Advances in treatment have led to a decrease in the rate of mother-to-child transmission of the HIV in high-income countries from 20% in 1994 to less than 1% for women receiving combination antiretroviral therapy in 2009.
Antiretroviral treatment is indicated for all pregnant women living with HIV-1 [1–3], because good control of viral load at delivery effectively prevents mother-to-child HIV transmission . Also, the risk of transmission decreases with the earlier such control is achieved . Methods for measuring viral load are increasingly sensitive, and this has led to the threshold for an undetectable viral load decreasing: currently in routine clinical practice in France a viral load of less than 50 copies/ml is considered to be the target for pregnant women, despite the absence of evidence that this is more effective than thresholds of 200 or 400 copies/ml for preventing mother-to-child transmission.
About 40% of HIV-infected pregnant women in France start cART during pregnancy. The viral load in one-third of these women does not fall to less than 50 copies/ml at delivery  and this is also true for 13–44% in other high-income countries [7–9]. The virological objective of 50 copies/ml may be impractical. Indeed, according to the current French guidelines, the virological objective for nonpregnant patients is a viral load less than 400 copies/ml within 3 months and less than 50 copies/ml within 6 months of cART initiation ; for pregnant women, an undetectable viral load (<50 copies/ml in practice) is recommended at delivery. The median duration of treatment at delivery for pregnant women commencing cART during pregnancy is only 3 months , such that the target of 50 copies/ml is therapeutically more demanding than for nonpregnant women.
Several studies have shown that the transmission rate decreases with longer duration of cART, and is particularly low if cART is started before pregnancy [5,10,11]. Consequently, cART has been initiated increasingly early during pregnancy in industrialized countries. This is inevitably associated with increasing duration of in-utero exposure of the fetus to antiretroviral drugs, and therefore potential toxicity [12–16].
Low pretherapeutic viral load, longer duration of cART and adherence to treatment are the main predictors of an undetectable viral load at delivery [8,9]. However, other pregnancy-related factors, due to physiological changes, may interfere with the response to cART initiated during pregnancy. A reduction of the plasma concentrations of protease inhibitors (lopinavir/r, saquinavir/r and atazanavir/r) has been described in pregnant women [8,17–20], although low concentrations of protease inhibitors administered during pregnancy at doses recommended for nonpregnant adults did not correlate with virological failure in one study .
To our knowledge, only two studies have compared the response to cART initiated during and outside pregnancy. The Swiss HIV Cohort study found a trend for a lower risk of virological failure (>1000 copies/ml) in pregnant than nonpregnant women . However, 59% of the pregnant women had started treatment before conception, and 75% of these women had an undetectable (<50 copies/ml) viral load before pregnancy. The second was conducted in a US cohort of women in which the evolution of viral load and CD4 cell count were similar when cART was initiated before or during their first pregnancy . Both studies were limited by the small numbers of women starting treatment during pregnancy (145 in the Swiss cohort, 70 in the US cohort).
Using data from three large French national cohorts including pregnant and nonpregnant women, we aimed to assess whether pregnancy may affect the early immunovirological response to cART, taking baseline immunovirological and sociodemographic characteristics into account.
We used data from three national multicenter prospective cohorts: the French Perinatal Cohort EPF (ANRS-CO-01/CO-11), COPANA (ANRS-CO-09) and PRIMO (ANRS-CO-06), all three coordinated by the same research team and involving the same network of clinical HIV specialists.
At the cutoff time for this analysis (2010), EPF had enrolled, since 1985, 16 182 pregnancies for 11 464 women infected by HIV (about 70% of deliveries of HIV-infected pregnant women in France). Since 1996, the PRIMO cohort had enrolled 1291 patients at the time of primary HIV infection. Between 2004 and 2008, the COPANA cohort had enrolled 800 patients with a recent (within the previous year) diagnosis of HIV-1, and who were ART-naive at baseline. The three cohort studies have been approved by the institutional review board of Cochin Hospital and the French computer database watchdog (Commission Nationale de l’Informatique et des Libertés).
Women who initiated cART before age 45 years were eligible for this study, provided that they were diagnosed after 2002 (to ensure access to similar therapeutic management and viral load measurement throughout the study period). cART was defined as an antiretroviral therapy comprising at least two nucleoside reverse transcriptase inhibitors combined with either a protease inhibitor (whether or not boosted) or a nonnucleoside reverse transcriptase inhibitor or a third nucleoside reverse transcriptase inhibitor. Women were excluded if they started cART within 6 months of the primary infection (n = 53), they started cART later than after 32 weeks of gestation (n = 69), since an undetectable viral load is difficult to attain after such short duration of treatments, often associated with late diagnosis and unfavorable environments, it was not possible to ascertain whether treatment was started during or outside pregnancy (date of last menstrual period or date of delivery missing, n = 11).
Overall, 779 women initiating cART while pregnant and 113 while nonpregnant were included in the analysis.
Definition of the outcomes
The primary outcome was virological efficacy according to two different thresholds (400 and 50 HIV RNA copies/ml), evaluated at M1 (1 month ± 15 days), M3 (3 months ± 45 days) and M6 (6 months ± 45 days) after cART initiation. The virological objective in clinical practice in several European countries is 50 copies/ml by delivery; the threshold of 400 copies/ml is recommended in the United States, and in France is the cutoff above which a programmed caesarian section is recommended [1,2]. A viral load decrease at least 2 log10 copies/ml during the first month, considered by French guidelines to be an indicator of the very early virological response, was also studied among women whose baseline viral load was above 5000 copies/ml. The CD4 count slope was investigated as a marker of the immunological response.
Baseline characteristics, virological efficacy and immunological response were compared between pregnant and nonpregnant women. Data for pregnant women were truncated at delivery.
At each time-point, the proportion of women with virological efficacy was estimated in the 818 (91.7%) patients whose baseline viral load was available. If viral load was undetectable, the viral load recorded was half of the value of the detection threshold. Complementary log–log multivariate regression models were built to estimate risk ratios for virological efficacy at M1, M3 and M6. Fractional polynomials were used to take the viral load (in log10 copies/mL) at cART initiation into account as accurately as possible; the model including the linear form of the baseline viral load was rejected in favor of the cube. Adjusted risk ratios (aRRs) were estimated in two nested models. The first aRR (aRR1) was adjusted for treatment duration and the cube of the baseline viral load and the second (aRR2), in addition, for baseline CD4 count, age, geographic origin, living in a couple and chronic hepatitis B. Adjustment for treatment duration aimed to avoid bias related to the right truncation due to delivery.
CD4 counts were plotted in LOWESS curves and square-root CD4 count slopes were compared for the first 3 months following cART initiation, by using random-effects models including random intercept and slope; models were adjusted for age, geographic origin, living in a couple, baseline viral load and AIDS status. The square-root of the CD4 count was used instead of CD4 count to provide a normal distribution; when plotting LOWESS curves and for mixed model prediction, the values were transformed back into CD4 counts to facilitate readability. The CD4 percentage increase was also modeled.
Two main analyses were done: one for all pregnant women initiating cART during pregnancy before 32 weeks, and one including only women having initiated cART during the first trimester (<15 weeks of gestation). Initiation of antiretroviral treatment early during pregnancy has been recommended until recently only for women who had an indication for their own health, so this sub-group of pregnant women was likely to be the most comparable to women initiating a cART outside pregnancy, especially in terms of baseline immunovirological status. In addition, more than 6 months of follow-up was available for most, allowing comparison of immunovirological response over 6 months.
Two sensitivity analyses were also performed. The first was restricted to women first treated with a protease inhibitor boosted by ritonavir (prescribed for almost all women initially during pregnancy). The second sensitivity analysis comprised only women starting cART in 2007 or later; all these women were likely to have had access to similar clinical care strategies.
All statistical tests were two-sided using a type I error of 0.05. Analyses were performed in intention-to-treat and computed using STATA (release 11; Stata Corp., College Station, Texas, USA) and SAS (release 9.2; SAS Institute Inc., Cary, North Carolina, USA).
The analysis included 892 women (779 pregnant and 113 nonpregnant). Socio-demographic characteristics, medical history, immunovirological status at cART initiation and first-line molecules are reported in Table 1. Pregnant women differed from nonpregnant women for several variables: they were younger, more often originated from sub-Saharan Africa, more often living in a couple, more often coinfected with hepatitis B virus, and more recently diagnosed for HIV (69.5% of pregnant women were diagnosed during pregnancy). At baseline, pregnant women had a better immunovirological status than nonpregnant women, and less often a history of AIDS. The baseline viral load was above 5 log10 copies/ml in 7.9% of pregnant versus 30.0% of nonpregnant women and the CD4 count was less than 200 cells/μl in 16.2 versus 36.6%, respectively. The women who initiated cART before 15 weeks of gestation (n = 78) were more comparable to nonpregnant women for baseline viral load (median: 4.3 and 4.4 log10 copies/ml) and CD4 count (median: 216 and 232 per μl) as they were likely to have started treatment for their own health. The first-line cART mostly included a ritonavir-boosted protease inhibitor regardless of gestational status (84.3% during and 61.1% outside pregnancy).
Virological response (400 copies/ml)
At M1, the proportion of women achieving a viral load less than 400 copies/ml was higher in the pregnant than nonpregnancy group (80.9 versus 61.5%, P = 0.001), as expected given their better immunovirological status at cART initiation (Table 2a). This association was not significant after adjustment for treatment duration and baseline viral load [aRR1: 1.3; 95% confidence interval (95%CI): (0.9–1.9)]. Further adjustment for baseline CD4 count, age, living in a couple, geographic origin and chronic hepatitis B did not modify the risk ratio [aRR2: 1.3; (0.8–2.0)]. The results for women initiating cART before 15 weeks of gestation were similar.
At M3, the virological response was high and comparable in pregnant (92.5%) and nonpregnant (88.1%, P = 0.17) women, and in women starting therapy before 15 weeks of gestation (86.9%). The adjusted RR was close to one [aRR2: 1.0; (0.7–1.4)].
Most pregnant women (74.0%) had delivered before M6 of treatment; those who had not were mostly women who had initiated cART before 15 weeks of gestation. All women in this group (100%) had a viral load less than 400 copies/ml at M6, with no significant difference for nonpregnant women at M6 (94.8%, P = 0.99).
Results were similar when the analysis was restricted to either women whose first cART comprised a protease inhibitor boosted by ritonavir (Table 2b) or those starting cART in 2007 or later (not shown).
Virological response (50 copies/ml)
Univariate analysis indicated that pregnant women were more likely to reach a viral load less than 50 copies/ml than nonpregnant women (35.6 versus 21.2%, P = 0.036) at M1, and this association disappeared after adjustment for baseline viral load [aRR1: 1.0; 95%CI: (0.6–1.9)] (Table 3a).
At M3, 61.5% of pregnant and 67.9% of nonpregnant women (P = 0.26) had a viral load less than 50 copies/ml. In multivariate models, starting cART during pregnancy was associated with a poorer virological response [aRR2: 0.6; (0.4–0.9)]. This association persisted when considering only women having initiated cART before 15 weeks of gestation [aRR2: 0.5; (0.2–0.9)].
At M6, 82.1% of pregnant women having initiated cART before 15 weeks of gestation and 87.0% of women having initiated cART whereas nonpregnant (P = 0.48) had a viral load less than 50 copies/ml. aRRs were of the same order of magnitude as at M3 [aRR2: 0.6; (0.3–1.1)].
These results did not change substantially when the analysis was performed for the subgroups whose first cART comprised a protease inhibitor boosted by ritonavir (Table 3b) or was initiated in 2007 or later (not shown).
Other virological outcomes
The proportion of women with a viral load decrease of at least 2 log10 copies/ml between cART initiation and M1 did not significantly differ according to gestational status: 65.1 during and 58.1% outside pregnancy (P = 0.38). This proportion was of the same magnitude (60.0%) in the subgroup of pregnant women beginning cART before 15 weeks of gestation.
Overall, the viral load closest to delivery was less than 400 copies/ml in 90.3% of pregnant women and 92.2% of women starting cART before 15 weeks. However, the threshold of 50 copies/ml was reached in only 63.8 and 67.2%, respectively. Nine infants were diagnosed as HIV-positive, giving a transmission rate of 1.2% (95%CI: 0.6–2.3).
Slopes of CD4 counts over the first 3 months of cART did not differ between pregnant and nonpregnant women (P = .14) (Fig. 1). This result was unchanged by adjustment for age, geographic origin, living in a couple, baseline viral load and AIDS status (P = 0.26), and also applied over 6 months of follow-up in women having started cART before 15 weeks of gestation (P = 0.78 in multivariate model). Analogous analyses with the percentage of CD4 instead of the number led to the same conclusion.
We found that when cART was initiated during pregnancy, nearly 90% of ART-naive women reached a viral load less than 400 copies/ml after 3 months and nearly 100% after 6 months of treatment; this is similar to what was observed for women starting when not pregnant. The vast majority of pregnant women (90.3%) achieved a viral load less than 400 copies/ml by delivery, consistent with other studies (82–93% [8,21]). CD4 recovery, assessed as either number or percentage of CD4 cells, was also similar in pregnant and nonpregnant women. The median increase was 99 cells/μl from cART initiation to delivery, very similar to that reported by Weinberg et al. (88 cells/μl) . These main findings suggest that pregnancy has no large effect on either the virological or the immunological responses to cART. This is consistent with the only two previous studies on this topic [16,17].
The proportion of pregnant women with a viral load less than 50 copies/ml was higher than that estimated in their US counterparts: 61.5 versus 42% at M3, 87.0 vs. 56% at M6 . Although the study designs and populations were different, this may nevertheless reflect the differences between US and French recommendations regarding the target viral load by delivery, resulting in more intensive therapy in France than in the United States.
However, multivariate models indicated a lower probability of a viral load less than 50 copies/ml among pregnant than nonpregnant women. This is unlikely to be due to different first-line cARTs, as a sensitivity analysis including only women receiving a boosted protease inhibitor gave the same results. This is also unlikely to have resulted from resistance to antiretroviral drugs, as all patients were ART-naive at baseline and the prevalence of genotypic resistance is generally low in African patients in France, a group constituting the majority of the pregnant women. Possibly, physiological factors, such as hormonal and immunological differences [23,24], could influence the response to cART during pregnancy. An increased volume of distribution of the drugs, hypoalbuminemia (partly due to hemodilution), vomiting and changes in hepatic drug metabolism and renal clearance may reduce the plasma concentrations of protease inhibitor [17–20].
As our analyses were performed in intention-to-treat, the effects on the virological response of physiological changes associated with pregnancy could have been partially offset by a better adherence during pregnancy [21,25,26] or differential treatment modifications. Nevertheless, therapeutic interruptions lasting more than 15 days were rare (3% for both pregnant and nonpregnant women). In addition, the first change of antiretroviral molecule occurred later in pregnant than in nonpregnant women, possibly a consequence of there being fewer alternate options during pregnancy. The reasons for change were not studied.
One of the main strengths of this study is that it was based on three multicenter cohorts with national coverage, coordinated by the same research team, with patients enrolled in the same network of hospitals managing HIV-infected patients. Also, many potential confounding factors were taken into account, including baseline viral load and CD4 count, sociodemographic factors (including age and geographic origin), and coinfection with hepatitis B.
Our analysis included seven-times more women having started cART during pregnancy than the two other studies addressing this issue [16,17], but the number of women who delivered after 6 months of therapy was nevertheless small. This is inevitable as few women initiate ART during the first trimester, consistent with French and international recommendations during the study period.
We conclude that pregnancy does not affect the response to cART, for following reasons: the proportions of pregnant and nonpregnant women with a viral load below 400 copies/ml were similar and high after 3 and 6 months; the immunological response, assessed as absolute numbers and percentages, was similar in the two groups; most women who failed to achieve a viral load less than 50 copies/ml by delivery had a viral load less than 400 copies/ml, and thus no indication for planned caesarian delivery and met the US virological objective.
Therefore, with current molecules and doses recommended in pregnancy, the main factor for not achieving an undetectable viral load by delivery may be the short duration of cART [5,10,11]. To decrease the viral load to less than 50 copies/ml, treatment could be started earlier, as previously suggested [5,8,9]; alternatively, the use of more intensive but shorter treatment could be considered. The optimal antiretroviral strategy needs to be identified, in part by risk–benefit analysis addressing the potential consequences for the fetus and the future child of high doses, other molecules and prolonged exposure in utero[12–16].
We thank all the Kremlin Bicêtre research team and the investigators of ANRS EPF, COPANA and PRIMO: CHU La Grave, Toulouse (Prevoteau Du Clary F., Viraben R.); CHU Pontchaillou, Rennes (Arvieux C., Bauville E., Chapplain JM., Cotten G., Dupont M., Dupré C., Fily F., Lassel L., Leroy H., Michelet C., Rasjtajack M., Revest M., Souala F., Tattevin P.); CHU, Brest (Ansart S., Bages-Jaffuel C., Duthé JC., Garre M., Jaffuel S., Le Moine P.); CHU, Nantes (Allavena C., Billaud E., Bonnet B., Bouchez S., Boutoille D., Brosseau D., Brunet C., Feuillebois N., Hue H., Mechinaud F., Mounoury O., Raffi F., Reliquet V., Sicot M., Talarmin JP.); CHU, Reims (Berger JL., Jaussaud R., Rouger C.); Centre Hospitalier André Mignot, Le Chesnay (Godin Collet C., Greder Belan A., Hentgen V., Messaoudi F., Ruquet M., Therby A.); Centre Hospitalier Bretagne Atlantique, Vannes (Cudeville C., Dos Santos A., Grellier A., Le Mabec B., Mouton Rioux V., Poinsignon Y.); Centre Hospitalier Bretagne Sud, Lorient (Moreau P., Niault M., Vaillant O.); Centre Hospitalier Départemental, La Roche sur Yon (Aubry O., Brossier JP., Esnault JL., Guimard T., Leantez-Nainville S., Leautez S., Perré P., Suaud I.); Centre Hospitalier Henri Duffaut, Avignon (Pichancourt G.); Centre Hospitalier Intercommunal de Cornouaille, Quimper (Bellot C., Blondin G., Duthé JC., Perfezou P.); Centre Hospitalier La Beauchée, Saint Brieuc (Beuscart C., Calvez C., Daniel C., Duhamel E., Gatel A., Le Moal S., Lemoal S., Pape E., Rohan J.); Centre Hospitalier Lyon Sud, Lyon (Massardier J.); Centre Hospitalier Régional Côte de Nacre, Caen (De La Blanchardiere A., Feret P., Martin A., Noyon V., Six M., Verdon R.); Centre Hospitalier William Morey, Chalon sur Saône (Florence M., Martha B., Martha S., Salle E.); Centre Hospitalier de Brive, (Abraham B., Lambert De Cursay G., Perino C., Pinet P.); Centre Hospitalier de Compiègne, (Domart Y., Merrien D.); Centre Hospitalier de Montluçon, (Antoniotti O.); Centre Hospitalier de Valence, (Champagne H., Riou R.); Centre Hospitalier de Vichy, (Regnier A.); Centre Hospitalier de la Haute Vallée de l’Oise, Noyon (Diab G., Grihon J.); Centre Hospitalier d’Alençon, (Houlbert D.); Centre Hospitalier d’Antibes, (Andreo S., Quinsat D., Tempesta S.); Centre Hospitalier, Vernon (Rouha M.); Centre Médical de l’Institut Pasteur, Paris (Consigny PH., Duvivier C., Touam F.); Centre hospitalier général, Aix en Provence (Allegre T., Blanc AP., Lafon C., Roubaud I.); Clinique du Blanc Mesnil*, (Balde P.); Hôpital Ambroise Paré, Boulogne (Reimann E., Rouveix E.); Hôpital Ambroise Paré, Boulogne* (Zenaty D.); Hôpital Ambroise Paré, Marseille (Chapus C., Philibert P.); Hôpital Américain, Reims (Berger JL., Graesslin O., Munzer M.); Hôpital André Rosemon, Cayenne (Elenga N.); Hôpital Antoine Béclère, Clamart (Benachi A., Bornarel D., Boué F., Chambrin V., Clech L., Fior R., Foix L’Hélias L., Labrune P., Levy A., Martinez V., Picone O., Raho M., Rannou MT., Schoen H., kansau I.); Hôpital Arnaud de Villeneuve, Montpellier (Benos P., Guigue N., Lalande M.); Hôpital Avicenne, Bobigny (Abgrall S., Barruet R., Bentata M., Bouchaud O., Dhote R., Honoré P., Lukiana T., Mechai F., Rouges F., Tuyet X.); Hôpital Beaujon, Clichy (Bensalah M., Ceccaldi-Carp P., Fantin B., Gellen-Dautremer J., Lefort A., Luton D., Pejoan H., Poder C., Tadlaoui A., Uludag A., Villemant-Uludag A., Zarrouk V.); Hôpital Bichat, Paris (Azria E., Bastian H., Bourgeois-Moine A., Bouvet E., Catalano G., Colasante U., Diallo B., El Alami Talbi N., Elaoun N., Faucher P., Fournier I., Godard C., Guiroy F., Huri V., Jadand C., Jestin C., Laurichesse JJ., Legac S., Leport C., Louni F., Matheron S., Pahlavan G., Phung B., Rajguru M., Ramani Z., Vivier V., Yazdanpanah Y., Yeni P., Zelie J.); Hôpital Boucicaut, Paris* (Lafay Pillet MC.); Hôpital Brabois, Vandoeuvre-les-Nancy (Boyer L., Briault A., May T., Neimann L., Wassoumbou S.); Hôpital Bretonneau, Tours (Bastides F., Bernard L., Besnier JM., Guadagnin P., Laplantine V., Nau P., Sajole G., Soufflet A., Sèvre C.); Hôpital Caremeau, Nîmes (Dendale-Nguyen J.); Hôpital Charles Nicolle, Rouen (Borsa-Lebas F., Brossard V., Caron F., Clavier B., Debab Y., Pinquier D., Pinto-Cardoso G.); Hôpital Civil, Strasbourg (Bernard-Henry C., Cheneau C., Cheneau M., De Mautort E., Entz-Werle N., Favreau J., Fischer P., Nisand I., Partisani M., Priester M., Rey D., Vayssière C., Weil M.); Hôpital Clémenceau, Caen (Beucher G., Brouard J., Goubin P.); Hôpital Cochin, Paris (Belarbi L., Boudjoudi N., Chakvetadze C., Firtion G., Fouchet M., Goupil I., Guillevin L., Launay O., Pannier E., Pietri MP., Salmon D., Salmon-Céron D., Silbermann B., Tahi T.); Hôpital Dupuytren, Limoges (Genet C., Weinbreck P.); Hôpital Dupuytren, Limoges* (De Lumley L.); Hôpital Edouard Herriot, Lyon (Chiarello P., Jeanblanc F., Livrozet JM., Makhloufi D., Touraine JL.); Hôpital Emile Muller, Mulhouse (Beck C., Beck-Wirth G., Benomar M., Drenou B., Halna JM., Michel C., Peter JM.); Hôpital Esquirol, Saint-Maurice* (Robin M.); Hôpital Européen Georges Pompidou, Paris (Buisson M., Jalbert C., Lucas ML., Manea M., Ptak M., Tisne-Dessus D., Weiss L.); Hôpital Foch, Suresnes (Bornarel D., Majerholc C., Zucman D.); Hôpital Foch, Suresnes* (Botto C.); Hôpital Font-Pré, Toulon (Allemand J., Assi A., Chéret A., Danielli E., Hittinger G., Huleux T., Lafeuillade A., Lambry V., Philip G., Rieu A.); Hôpital Francilien Sud, Evry-Corbeil (Blasquez G., Chevojon P., Devidas A., Farvacque I., Granier M., Guignier M., Lakhdari Y., Marchand C., May A., Nguyen R., Turpault I.); Hôpital François Quesnay, Mantes La Jolie (Delanete A., Doumet A., Furioli J., Granier F., Perrone V., Salomon JL.); Hôpital Gabriel-Montpied, Clermont Ferrand (Beytout J., Gourdon F., Jacomet C., Tas H.); Hôpital Gui de Chauliac, Montpellier (Atoui N., Baillat V., Crisol C., Lemoing V., Makinson A., Marc Jacquet JM., Merle De Boever C., Psomas C., Reynes J., Tramoni C., Vidal M., Villadero A.); Hôpital Gustave Dron, Tourcoing (Aissi E., Ajana F., Alcaraz I., Allienne C., Aïssi E., Baclet V., Bonne S., Chéret A., De La Tribonniere X., Huleux T., Melliez H., Meybeck A., Riff B., Senneville E., Valette M., Viget N., Yazdanpanah Y.); Hôpital Haut Lévèque, Bordeaux (Chaigne De Lalande S., Greib C., Lazaro E., Pellegrin JL., Raymond I., Viallard JF., Wirth G.); Hôpital Henri Mondor, Créteil (Dominguez S., Dumont C., Jung C., Lascaux AS., Lelievre JD., Levy Y.); Hôpital Hôtel Dieu, Paris (Compagnucci A., Cros A., Gilquin J., Hadacek B., Huchon G., Maignan A., Sobel A., Thu-Huyn N., Viard JP., Zak-Dit-Zbar O.); Hôpital Intercommunal André Grégoire, Montreuil (Fried D., Heller-Roussin B., Riehl C., Winter C.); Hôpital Intercommunal, Créteil (De Lacroix Szmania I., Elharrar B., Garrait V., Haddad B., Ledudal P., Lemerle S., Pichon C., Richier C., Richier L., Touboul C.); Hôpital Intercommunal, Evreux (Allouche C., Johnson A., Touré K.); Hôpital Intercommunal, Montfermeil (Dehlinger M., Echard M., Mullard C., Talon P.); Hôpital Jean Rostand, Ivry* (Jault T., Jrad I. ); Hôpital Jean Verdier, Bondy (Benoist L., Bolie S., Coupard M., Delonnoy C., Fein O., Jeantils V., Lachassine E., Moreou L., Picard F., Prendki V., Stirnemann J., Tassi S.); Hôpital Jeanne de Flandres, Lille (Ajana F., D’angelo S.); Hôpital Joseph Ducuing, Toulouse (Garipuy D., Saint-Dizier F.); Hôpital La Timone, Marseille (Thuret I.); Hôpital Lariboisière, Paris (Ayral D., Bergmann JF., Ciraru-Vigneron N., Diemer M., Durel A., Mouchnino G., Parrinello M., Rami A.); Hôpital Les Abymes, Pointe à Pitre (Bataille H., Beaucaire G., Lamaury I.); Hôpital Louis Domergue, La Trinité* (Hugon N.); Hôpital Louis Mourier, Colombes (Bloch M., Cahitte I., Crenn-Hebert C., Duro D., Floch-Tudal C., Mahe I., Manceron V., Mandelbrot L., Mazy F., Meier C., Montoya B., Mortier E., Simonpoli AM., Zeng A.); Hôpital L’Archet, Nice (Bongain A., Cassuto JP., Cua E., Dellamonica P., Deville A., Durant J., Fuzibet JG., Galiba E., Leplatois A., Mondain –Miton V., Monpoux F., Naqvi A., Publiese P., Quaranta M., Rosenthal E., Serini MA., Sohn C., Vassallo M.); Hôpital Marc Jacquet, Melun (Boussard JL., Chardon P., Le Lorier B., Pauly-Ravelly I.); Hôpital Marechal Joffre, Perpignan (Aumaitre H., Bachelard G., Eden O., Ferreyra M., Malet M., Medus M., Neuville S., Saada M.); Hôpital Minjoz, Besançon (Gil H.); Hôpital Necker, Paris (Avettand-Fenoël V., Ayroles C., Blanche S., Cayol V., Duvivier C., Frange P., Le Mercier D., Mourey MC., Rouzioux C., Touam F., Verber F.); Hôpital Nord, Amiens (Decaux N., Douadi Y., El Samad Y., Gondry J., Li Thiao Te V., Pautard B., Schmit JL.); Hôpital Nord, Marseille (Mokhtari S., Moreau J., Schlossers M., Van Der Gheynst E.); Hôpital Nord, Saint Etienne (Billiemaz K., Fresard A., Ronat V.); Hôpital Notre Dame du Bon Secours, Paris (Aufrant C.); Hôpital Orléans, Orléans (Arsac P.); Hôpital Paris La Roseraie, Aubervilliers* (Rozan MA.); Hôpital Pasteur, Colmar (Audhuy B., Blaison G., Martinot M., Mohseni Zadeh M., Mothes A., Pachart A., Plaisance N., Richard A.); Hôpital Paul Brousse, Villejuif* (Dussaix E.); Hôpital Paule de Viguier, Toulouse (Antras M., Armand E., Berrebi A., Tricoire J.); Hôpital Pellegrin, Bordeaux (Blanchard E., Cazanave C., Dauchy F., Douard D., Dupon M., Dutronc H., Lacaze-Buzy L., Merlet A., Neau D., Pistone T., Ragnaud JM., Raymond I., Roux D., Schaeffer V.); Hôpital Pierre Zobda-Quitman, Fort-de-France (Abel S., Avenin G., Baringthon C., Beaujolais V., Cabié A., Hochedez P., Hurtrel G., Liautaud B., Pierre-Francois S., Vignier N.); Hôpital Pitié-Salpétrière, Paris (Adeb N., Amirat N., Ben Abdallah S., Bonmarchand M., Bourse P., Bricaire F., Caby F., Edeb N., Herson S., Iguertsira M., Katlama C., Lambert-Niclot S., Laubies A., Nafissa S., Naime-Alix AF., Pauchard M., Pichon F., Quetin F., Remidi H., Schneider L., Seang S., Simon A., Sommer J., Stitou H., Tubiana R., Valantin MA., Vourch JL.); Hôpital Purpan, Toulouse (Balzarin F., Chauveau M., Marchou B., Obadia M.); Hôpital Raymond Poincaré, Garches (Berthe H., Carlier R., De Truchis P., Perronne C.); Hôpital René Dubos, Pontoise (Coursol A., Youssef M.); Hôpital Robert Ballanger, Aulnays (Favret V., Gerbe J., Goldenstein C., Questiaux E., Zakaria A.); Hôpital Robert Bisson, Lisieux (Geffray L.); Hôpital Robert Debré, Paris (Bellaton E., Borie C., Boumediene MA., De Lauzanne A., Faye A., Garion D., Leveille S., Levine M., Matheron S., Ottenwalter A., Tadlaoui A.); Hôpital Saint André, Bordeaux (Aslan C., Bernard N., Bonarek M., Bonnet F., Caldato S., Delaune J., Hessamfar M., Lacoste D., Louis I., Malvy D., Morlat P., Paccalin F., Vandenhende MA.); Hôpital Saint Claude, Basse-Terre* (Sibille G.); Hôpital Saint Jacques, Besançon (Bourdeaux C., Chirouze C., Drobacheff-Thiebaut C., Estavoyer JM., Faucher JF., Foltzer A., Hoen B., Maillet R., Parcelier F., Thiebaut-Drobacheff MC.); Hôpital Saint Louis, Paris (Aslan A., Balkan S., Clavel F., Colin De Verdiere N., De Castro N., Delgado J., Ferret S., Gallien S., Garrait V., Gatey C., Gerard L., Goguel J., Gérard L., Lafaurie M., Lambert F., Lascoux-Combe C., Loze B., Molina JM., Morel P., Oksenhendler E., Parlier S., Pavie J., Pintado C., Ponscarme D., Rozenbaum W., Saint-Marc T., Sereni D., Séréni D., Taulera O., Timsit J., Van Dien L.); Hôpital Saint Michel, Paris* (Aufrant C.); Hôpital Saint-Antoine, Paris (Bollens D., Bui-Pha E., Campa P., Carbonne B., Desplanque N., Girard PM., Krain A., Lefebvre B., Lupin C., Meyohas MC., Picard O., Rodriguez J., Tredup J., Valin N.); Hôpital Sainte Marguerite, Marseille (Debreux C., Drogoul MP., Fabre G., Gastaut JA., Ivanova A., Poizot-Martin I., Zaegel-Faucher O.); Hôpital Sainte Musse, Toulon (Hittinger G., Philip G.); Hôpital Simone Veil, Eaubonne (Koza MA., Lepretre A.); Hôpital Tenon, Paris (Bani Sadr F., Bastard JP., Berrebi V., Bonnard P., Capeau J., Chakvetadze C., Fellahi S., Hervé F., Lebrette MG., Lyavanc T., L’Yavanc T., Pialoux G., Selleret L., Slama L., Vigouroux C.); Hôpital Trousseau, Paris (Benifla JL., De Montgolfier I., Dollfus C., Tabone MD., Vaudre G.); Hôpital Victor Dupouy, Argenteuil (Allisy C., Brault D., Genet P., Gerbe J., Sutton L.); Hôpital Victor Fouche, Fort-de-France (Hatchuel Y., William C. ); Hôpital de Bastia, (Pincemaille O.); Hôpital de Bicêtre, Le Kremlin Bicêtre (Blanc A., Bolliot C., Bourdic K., Chaix F., Colmant C., Delfraissy JF., Dulucq MJ., Duracinsky M., Escaut L., Fourcade C., Fourn E., Fuchs F., Ghosn J., Goujard C., Hammou Y., Jennequin C., Jrad I., Mazingue F., Peretti D., Quertainmont Y., Rannou MT., Segeral O., Seguin V., Vittecoq D.); Hôpital de Compiègne*, (Lagrue A.); Hôpital de Creil, (Carpentier B., Duval-Arnould M., Kingue-Ekollo C.); Hôpital de Dourdan*, (Ercoli V.); Hôpital de Dreux*, (Denavit MF.); Hôpital de Fontainebleau, (Al Aissa K., Dallot MC., Routier C.); Hôpital de Gonesse*, (Balde P.); Hôpital de La Seyne sur Mer, (Chamouilli JM.); Hôpital de Lagny, (Algava G., Chalvon Demersay A., Dagot B., Froguel E., Gourdel B., Simon P., Tassi S.); Hôpital de Longjumeau, (Abbara A., Bailly-Salin P., Chaix F., De San Pedro C., Ducrocq S., Seaume H., Turpault I.); Hôpital de Lôches, (Girard JJ., Simonet V. ); Hôpital de Meaux, (Karaoui L., Lefèvre V. ); Hôpital de Meulan*, (Seguy D.); Hôpital de Nanterre*, (Karoubi P.); Hôpital de Nanterre, (Daneluzzi V., Gerbe J., Ruel M.); Hôpital de Neuilly sur Seine*, (Berterottiere D.); Hôpital de Poissy-Saint-Germain en Laye, (Couderc S., Rousset MC.); Hôpital de Saint Martin, (Bissuel F., Walter V.); Hôpital de Saint-Denis, (Allemon MC., Bolot P., Dandris S., Ekoukou D., Ghibaudo N., Khuong-Josses MA.); Hôpital de Sèvres*, (Segard L.); Hôpital de Villeneuve Saint Georges, (Bantsimba J., Caraux Paz P., Cazenave B., Chacé A., Dellion S., Guillot F ., Matheron I., Patey O., Richier L., Tilouche S.); Hôpital de la Conception, Marseille (Cravello L., Gallais H., Ravaux I., Stein A., Valon A.); Hôpital de la Croix Rousse, Lyon (Augustin-Normand C., Baudry T., Brochier C., Cotte L., Koffi J., Labaune JM., Lafon A., Le Thi T., Miailhes P., Pailhes S., Peyramond D., Roussouly MJ., Schlienger I., Tariel O., Thoirain V., Trepo C.); Hôpital de la Côte Basque, Bayonne (Bonnal F., Cayla C.); Hôpital des Feugrais, Elbeuf* (Lahsinat K.); Hôpital des Métallurgistes*, Paris (Rami M.); Hôpital du Bocage, Dijon (Briandet C., Buisson M., Piroth L., Treuvelot S.); Hôpital du Lamentin, Le Lamentin* (Monlouis M.); Hôpital d’Aix en Provence*, (Tadrist B.); Hôpital d’Angers, (Abgueguen P., Chennebault JM., Fialaire P., Fournié A., Loison J., Pichard E., Rabier V., Rehaiem S.); Hôpital d’Orsay, (Chanzy S., De Gennes C., Isart V.); Hôpital d’enfants, Dijon (Martha S., Reynaud I.); INSERM U1018, Villejuif and Le Kremlin-Bicêtre (Boufassa F., Boumaza B., Charles MA., Dray-Spira R., Essabbani A., Legeai C., Meyer L., Seng R.); INSERM U897, Bordeaux (Thiébaut R.); Institut Mutualiste Montsouris, Paris (Bodard L., Gibert S., Meudec A., Sauret I.); Institut Mutualiste Montsouris, Paris* (Carlus Moncomble C.); Institut d’Hématologie-Oncologie Pédiatrique, Lyon (Bertrand Y., Kebaïli K., Tache N.); Maternité Régionale A. Pinard, Nancy (Hubert C.); Médecin généraliste, Paris (Miodovski C.)
(*sites EPF closed)
Funding: This work was supported by the Agence Nationale de Recherche sur le Sida et les hépatites virales (ANRS) and the Fondation pour la Recherche Médicale.
Conflicts of interest
J.L.C., C.L., J.P.T., L.M.: No conflict of interest. A.R.: Master scholarship from Fondation pour la recherche médicale; J.W.: Her institution received support for travel to meetings from ANRS-INSERM and grants from ANRS-INSERM, Abbott, Iatex, ViiV Healthcare, Parexel. C.G.: expert testimony (Gilead, Janssen), payment for development of educational presentations from Gilead. C.R.: Her institution received grant from ANRS, patents and royalties from HIV. L Mandelbrot: lecturing honorariae from BMS, Gilead and MSD. R.T.: payment for board membership by Gilead and lecturing honorariae from BMS and Gilead. Payment for development of educational presentations by University Paris 6. Money received by his institution from MSD for travel/accommodations / meeting expenses.