Continuous Spread of HIV-1 Subtypes D and CRF01_AE in France from 2003 to 2009

Abstract

Among 61 and 35 patients who were infected in France by viruses of the rare clades D and CRF01_AE, respectively, approximately half of them originated from areas where HIV-1 is endemic, but the data showed that both clades have spread in the French indigenous population, particularly in men having sex with men (MSM).

TEXT

One of the characteristics of the human immunodeficiency virus type 1 (HIV-1) pandemic is the tremendous genetic heterogeneity of the virus, leading to the identification of many subepidemics due to specific variants. Group M viruses, the pandemic branch of HIV-1, can be subdivided into phylogenetic subtypes (named A to D, F to H, J, and K), sub-subtypes (A1 and A2, F1 and F2), and circulating recombinant forms (CRFs). The most prevalent HIV-1 genetic forms worldwide are subtypes A, B, and C (7, 12). However, there is a specific geographic distribution pattern for HIV-1 subtypes, and less prevalent genetic forms can reach high proportions locally (2, 9, 12, 18, 20, 24). The relevance of CRFs in the global HIV-1 pandemic has also become clear, and particular CRFs are the predominant forms in Southeast Asia (CRF01_AE) and in West Africa (CRF02_AG) (12, 23).

Subtype B is the most prevalent genetic form in Western Europe. However, in various countries with large immigrant communities, the epidemiology of HIV-1 infection has evolved over the past decade toward a marked increase in the circulation of non-B strains (1, 17, 26, 29). The recent report of HIV-1 CRF01_AE outbreaks among injection drug users in Sweden and Finland demonstrated that standing social networks with risky behavior can contribute to the spread of HIV-1 subtypes otherwise barely detected in Western Europe (27). In France, HIV transmission disproportionately affects certain risk groups and seems to be out of control in the population of men having sex with men (MSM), who make up 48% of all new infections and have a persistently high incidence of 1% per year (16).

To assess the circulation of rare HIV-1 subtypes in France, we studied the epidemiological characteristics of patients infected with HIV-1 subtype D and CRF01_AE identified through the national HIV cases surveillance program from January 2003 to December 2009. We decided to focus on HIV-1 group M clades D and CRF01_AE for various reasons. First, these two genetic forms have different epidemiological characteristics. The HIV-1 subtype D prevalence in Africa is slowly decaying, and its introduction in France is supposed to rely mainly on Sub-Saharan African immigration. Conversely, the CFR01_AE epidemic has been highly dynamic in Southeast Asia (7, 12). Second, it has been suggested that HIV-1 clades D and CRF01_AE are associated with faster disease progression (3, 13, 19). Third, the serotyping assay used as a first rapid screening tool for the surveillance of HIV diversity in France discriminates efficiently the subtypes B, D, and CRF01_AE infections compared to other subtypes (4, 5, 21).

The mandatory anonymous HIV case reporting was implemented in France in 2003 (16, 26). It is linked to a virological surveillance program that was approved by the French National Ethics Committee (CNE) and the French computer database watchdog commission (CNIL). When a newly diagnosed HIV case is notified, the laboratory is asked to send dried serum spots (DSS) from the stored serum sample to the HIV French National Reference Center (NRC). The type, group, and subtype of HIV are determined using a subtype-specific enzyme-linked immunosorbent assay (SS-ELISA) applied to serum eluted from DSS (4, 5), shown previously as presenting good predictive values for HIV-1 subtypes D and CRF01_AE (21). Therefore, all new HIV-1 diagnoses notified from January 2003 to December 2009 that scored positive for these two HIV-1 genetic forms by serotyping were selected and further analyzed by genotyping (Fig. 1). The genotyping analyses were conducted on a 415-bp fragment of the envelope glycoprotein gp41 coding sequence amplified by reverse transcription-PCR (RT-PCR) from RNA extracted from DSS, as described previously (6, 22, 28). Additional D and CRF01_AE infection cases were detected through other independent genotyping analyses conducted to complement the serotyping approach. Indeed the serotyping assay can be interpreted only if infected patients have developed enough antibodies when they are tested. As a consequence, most serotyping failures are observed in patients with recent HIV-1 infections. These complementary analyses were conducted on DSS collected from January 2003 to January 2005 for which a nontypeable result by SS-ELISA was obtained and on DSS collected during the first semester of years 2007, 2008, and 2009 from recently infected patients and for which a nontypeable or a non-B result was obtained (Fig. 1).

Fig 1.

Algorithm used to select the samples that were analyzed by RT-PCR followed by sequencing and phylogenetic analysis. SSEIA, subtype-specific enzyme-linked immunosorbent assay.

Between January 2003 and December 2009, 26,043 new HIV diagnoses were reported to the InVS for which a DSS was available. HIV-1 group M-positive serotyping results were obtained for 23,603 individuals. The SS-ELISA yielded positive results for subtypes D and CRF01_AE in 446 and 131 DSS, respectively (Fig. 1). RNA was extracted from these 577 DSS and subjected to RT-PCR. Amplified DNA products were obtained for 176 samples (31%), including 140 from patients with a SS-ELISA-defined subtype D infection and 36 from patients with a SS-ELISA-defined CRF01_AE infection. Our previous experiences suggest that the failure to obtain a RT-PCR product could be attributed mainly to a viral load below 10,000 copies/ml in those samples (22, 26). Among the 140 patients with a SS-ELISA-defined subtype D infection, 54 (39%) were confirmed as subtype D. Of the others, there were 31 subtype B, 4 subtype A, 3 subtype G, 2 subtype C, 25 CRF02_AG, 5 CRF06_cpx, 3 CRF01_AE, 2 CRF11_cpx, and 11 unspecified non-B infections. A high proportion of genotype B infections (22%) was detected among the SS-ELISA-defined subtype D infections. This result can be explained by the fact that B and D subtypes are considered two sub-subtypes, with subtype D as the early-B African genetic form (7). Among the 36 patients with a SS-ELISA-defined CRF01_AE infection, 16 (44%) were confirmed as CRF01_AG, 6 were subtype B, 2 subtype C, 1 subtype A, 1 subtype D, 1 subtype G, 4 CRF02_AG, and 5 unspecified non-B infections. At this stage, we can conclude that the specificity of the SS-ELISA to predict for subtypes D or CRF01_AE was lower than estimated previously. It can be explained by the small size and the restricted geographic origins of the populations used at that time (21).

The genotyping analyses conducted in complementation of the serotyping approach allowed the detection of 6 and 16 additional patients infected by HIV-1 subtypes D and CRF01_AE, respectively (Fig. 1). Overall, we identified 61 patients with genotype-confirmed HIV-1 subtype D infection and 35 patients with genotype-confirmed HIV-1 CRF01_AE infections. These identifications were scattered over the study period of 7 years, with 4 to 10 cases of HIV-1 subtype D infection and 2 to 10 cases of CRF01_AE infection detected per year, including 4 and 3 primary infections with subtype D and CRF01_AE, respectively (Table 1).

Table 1.

Characteristics of patients with genotype-confirmed clade D or CRF01_AE infections

Characteristic	No. of cases
	HIV-1 subtype D	HIV-1 CRF01_AE
Yr (no. of new HIV diagnoses with DSS)
2003 (2,821)	6	4
2004 (3,395)	10	6
2005 (3,777)	12	4
2006 (3,584)	11	2
2007 (3,403)	4	5
2008 (3,379)	12	4
2009 (3,244)	6	10
Total (23,603)	61	35
Sex and transmission category
Heterosexual male	9	9
MSM	14	8
Other/unknown male	9	7
Heterosexual female	25	8
Other/unknown female	4	3
Age group
<30 yr	5	5
30–39 yr	12	12
40–49 yr	24	5
≥50 yr	20	13
Country of birth
France	18	8
	10 MSM	5 MSM
	6 heterosexuals	3 heterosexuals
	2 unknown
Other European country	3	2
Sub-Saharan Africa	28	7
	24 heterosexuals	7 heterosexuals
	1 MSM
	3 unknown
North Africa	1
America	3
Asia		6
		2 heterosexuals
		2 MSM
		2 other/unknown
Unknown	8	12
Region of residence
Paris and surrounding area	32	16
Other	22	16
Unknown	7	3
Clinical stage at presentation
Primary infection	4	3
Asymptomatic	25	15
Symptomatic without AIDS	9	2
AIDS	13	7
Unknown	10	8

Among the 61 patients infected with HIV-1 subtype D, 28 of 53 with a known place of birth (53%) came from Sub-Saharan Africa, whereas only 18 (34%) were French natives (Table 1). The Sub-Saharan Africans (n = 28) originated from Congo (n = 12), the Democratic Republic of Congo (DRC; n = 5), Angola (n = 3), Cameroon (n = 2), Mali (n = 2), Côte d'Ivoire (n = 2), Central African Republic (n = 1), and Gabon (n = 1). Most of these Sub-Saharan Africans (24 out of 28) declared to be heterosexuals. Conversely, a majority of subtype D-infected patients born in France were MSM (10 out of 18). The existence of clusters was ascertained based on high bootstrap values (≥99%) and short branch lengths (genetic distance of <0.02%) (Fig. 2). Three clusters were observed. One cluster involved three men living in the same region for which HIV diagnoses were notified in 2008 (n = 2) and 2009 (n = 1). Two of them were MSM, and the transmission category of the third one was unknown (Fig. 2A, large gray arrow). All three were identified as infected for less than 180 days using an enzyme immunoassay for recent infection (data not shown), and the complementary clinical data provided by the physician in the notification for the two MSM stated that they had signs of primary infection. It is also of interest to note that two other primary infections that were declared among patients infected by subtype D viruses occurred in MSM, one in 2007 and one in 2008. The two other clusters concerned patients probably infected through heterosexual contacts (Fig. 2A, thin black arrows). One cluster involved one man and one woman, both originating from Congo and living in Paris. The second cluster involved two women, one originating from Congo and the other originating from DRC, both living in the Paris suburb area.

Fig 2.

Phylogenetic trees of sequences from HIV-1 subtype D-infected patients (A) and CRF01_AE-infected patients (B). The 415-bp gp41 coding sequences used for the genotyping were analyzed comparatively using the MEGA analysis program with the neighbor-joining method. Distances were calculated using the Kimura two-parameter matrix (transitions/transversions ratio = 2.0). Bootstrap analysis (500 replications) was used to test the reliability of the branching order. Only bootstrap values equal or superior to 50 are indicated. The trees were rooted using the subtype A sequence A.2002.KTB035 and the group O sequence O.SN.99.AJ302646. The legend for the symbols is shown at the bottom of the figure (IDU, intravenous drug use). The scale bar at the bottom of the figure indicates the number of base substitutions per site. (A) The analysis was conducted on 57 sequences out of 61. The four samples eliminated were shorter sequences for which we encountered sequencing difficulties. Nine subtype D reference sequences selected among the HIV sequence database (http://hiv-web.lanl.gov) were included in the analysis (D.FJ388945, D.EU853116, D.DQ054367, D.U88824, D.AF457090, D.AY253311, D.AY795907, D.AY371157, D.K03454). The large gray arrow shows the cluster that involved MSM, as described in the text. The thin black arrows show the clusters that involved patients infected through heterosexual contacts. (B) The analysis was conducted on 33 sequences out of 35. The two samples eliminated were shorter sequences for which we encountered sequencing difficulties. Thirteen CRF01_AE reference sequences (ten from Asian isolates and three from African isolates) selected from the HIV sequence database (http://hiv-web.lanl.gov) were included in the analysis (AE.CN.DQ859180, AE.HK.DQ234790, AE.CN.GQ845125, AE.CN.FJ441290, AE.TH.AY945722, AE.U54771, AE.TH.DQ314732, AE.TH.AY358062, AE.TH.AY945719, AE.TH.AB220944, AE.CF.U51188, AE.CF.AF197340, AE.CF.AF197341).

Among the 35 patients infected with CRF01_AE variants, 7 (30%) and 6 (26%) of 23 with a known place of birth came from Sub-Saharan Africa and Asia, respectively, whereas 8 (35%) were French natives (Table 1). The Sub-Saharan Africans originated from various countries of Central Africa (2 from Cameroon, 2 from Central African Republic, 1 from Congo, 1 from DRC, and 1 from Togo), and all of them declared to be heterosexuals. The patients born in Asia (n = 6) came from Thailand (n = 3), China (n = 2), and Vietnam (n = 1), two of them being MSM. Among the 8 patients born in France, 5 were MSM, 4 having been diagnosed during the 2007 to 2009 period. There was no evidence of cluster (Fig. 2B).

An increasing proportion of new HIV diagnoses in various European countries is attributable to infections with viruses of non-B subtypes. In France, nearly half of the newly diagnosed HIV-1 infections between 2003 and 2009 was associated with non-B subtypes, whereas non-B viruses were responsible for only 15% of new cases in 1996 to 1998 (8, 10, 26). Several recent studies indicate that approximately 50% of these HIV-1 non-B infections in France were due to CRF02_AG variants, which is the predominant genetic form in West Africa (11, 26). However, the spread of other less frequent subtypes remains difficult to assess without screening large numbers of samples. Here, we report data on infections involving HIV-1 clades D and CRF01_AE obtained over a period of 7 years. Our interest for these two clades were mainly the differences observed in their global epidemiological characteristics, which make them probes to assess the spread of rare HIV-1 strains in France. We also focused on these clades because of their possible (CRF01_AE) or probable (D) association with a faster disease progression (3, 13, 19). Furthermore, a recent report suggested a possible higher infectivity of subtype D variants in the mother-to-child transmission setting (14). In addition, it was shown that the genotypic determination of HIV-1 tropism with algorithms built for serotype B were poorly specific for predicting CXCR4 use by subtype D viruses and would necessitate specific rules (25).

Approximately half of the patients who were infected by HIV-1 subtype D (53%) or CRF01_AE variants (56%) originated from countries where these viruses are endemic. However, our data demonstrated that both HIV-1 clades D and CRF01_AE have spread in the French indigenous population, particularly in MSM. The data did not suggest an increasing prevalence over the study period but clearly showed a continuous spread since several primary infections due to both subtypes were identified. The presence of rare non-B variants in MSM, including those recently infected, suggests a sustained genetic diversity of HIV strains within the MSM community with the possibility to generate new recombinant forms when superinfections occur (15). The phylogenetic evidence for a cluster of a subtype D strain among MSM might be an early indicator for a possible outbreak involving this rare HIV-1 subtype. Due to both the suggested greater pathogenicity of this subtype and the high incidence of HIV-1 in French MSM (16), it remains particularly necessary to continue to carefully monitor HIV-1 diversity, particularly in populations with high-risk behavior.