Analysis of women diagnosed with HIV-1 in Israel revealed that most were not born in Israel. In 2010–2012, 44.7% were immigrants from SSA and 31.3% were from the FSU. In more recent years (2016–2018), 50% were from the FSU, while only 27.7% originated from SSA (p < 0.001). The most prevalent viral subtype, changed accordingly, from subtype C, characterizing HIV-1 in SSA, in 2010–2012, to subtype A, characterizing FSU, in 2016–2018 (p < 0.014). These results are in concordance with the waves of immigration from SSA and Eastern Europe to Israel in 2010–2018. A similar increase in the prevalence of subtype A carriers was recently reported in Germany and in other west-European countries, due to an increased flow of refugees, mainly from the FSU, into Europe, and especially into Germany (19).
The low CD4 counts, median of 263 cells/mm*3 noted in here for women, suggest late diagnosis. Moreover, women from SSA, as well as those who immigrated from the FSU, had significantly lower CD4 counts at diagnosis compared to Israeli-born women. Missed opportunities for early diagnosis were already suggested for at least 33% of the Israeli HIV population (20). Late diagnosis was also recently reported to characterize over a half of the women diagnosed in Europe in 2018 (21). Our data corroborate these results and highlight the need for improved HIV policies targeting new female immigrants. These can include offering HIV testing to all women immigrating from concentrated and generalized HIV epidemic regions, such as the FSU and SSA, respectively, be soon after their arrival. Also, as most of the women are diagnosed at the reproductive age (median age of diagnosis was 38 years), universal testing for HIV-1 infection during pregnancy should be employed, without limiting it to a selected group, e.g., immigrants from SSA, as currently performed (22). It was already demonstrated that a universal approach to perinatal HIV testing achieves the best health outcomes and is cost-effective across a range of HIV-1 prevalence settings (23).
TRDM were identified in 10.4% of women diagnosed in 2010–2018. Prevalence of women with NNRTI, NRTI and PI TDRM was 7.1%, 3% and 1.8%, respectively. The proportion of women diagnosed with any TDRM and especially with NNRTI TDRM increased significantly in more recent years, reaching 14.4% (or 13.3% for women with NNRTI TDRM) of all women diagnosed in 2016–2018. In a recent analysis of HIV diagnoses in 2017 in 9 European countries, the overall prevalence of resistance mutations in treatment naïve patients was 13.5% and that of NNRTI was 7.7% (24). Although these results are similar to our findings in women, they are likely an overestimation of the actual TDRM rate in Europe, as all resistance mutations included in the Stanford HIVdb were considered (16, 17). In general, changes in prescribing practices over time, the high genetic barrier of PI and the lower genetic barrier of NNRTIs, most likely explain the changing rates of drug class-related TDRM (25). However, the overall high rate of resistance mutations, the ongoing increase in transmission of resistant viruses especially in more recent years and the high rate of individuals on antiviral therapy worldwide, mandates continuous monitoring of pretreatment resistance mutations in Israel and around the world.
NNRTI are not considered the preferred first line therapies, however they are still included in at least some of these regimens (5, 25). In the current study, K103N/S (which confers high-level or intermediate cross-resistance to the NNRTIs efavirenz, nevirapine and delavirdine) was the most prominent NNRTI TDRM (4.2%) and more prevalent in HIV-1 subtype B carriers, as previously reported (26). As current guidelines permit the use of efavirenz among women of childbearing potential, this rather frequent TDRM should not be disregarded. The polymorphic E138 was the most frequently mutated NNRTI site. This naturally occurring polymorphism that blocks the NNRTI-binding pocket, is known to affect rilpivirine binding and may cause lower susceptibility to this drug (27). A systematic review that assessed the prevalence of rilpivirine-related TDRMs in 65 countries already reported an association between E138 mutations and HIV-1 subtypes C (6.1%), and A (3.3%) (28). In the current study, it was identified in 5.6% of all women, irrespective of the viral subtype. As rilpivirine-based dual therapy is still considered a legitimate treatment option, resistance testing in all patients prior to rilpivirine therapy should be employed. The most prominent NRTI accessory non-polymorphic mutated site was A62V (5.6% prevalence), which influences replication fidelity and viral fitness in the context of multi-drug resistance mutations (17). A62V, which was reported to be widespread in subtype A viruses in the FSU (17), was also significantly more prominent in HIV-1 subtype A in our analysis. However, A62V does not interfere with current therapy.
While there was no significant difference between overall TDRM rates in women originating from different countries, significantly higher NNRTI TDRM rates (16.3%) were identified in women born in Israel compared to those born in SSA (8.9%) or FSU (4.9%, p = 0.014). Previously, in a study that assessed patients diagnosed between 1999–2003 in Israel, resistance mutations were reported in 14.8% of newly-diagnosed, treatment-naïve HIV patients, 28.6% of whom were known to have been infected in Israel (7). Together these results suggest continuous ongoing local circulation of drug-resistant viruses. An in-depth characterization of all HIV-1 patients identified in 2010–2018 is ongoing.
Our study has several limitations. The main inherent limitation is the overall small number of women positive for HIV diagnosed in Israel. Also, resistance analysis was not performed for all women. However, a representative sample, selected each year by stratified selection design, was used for sequencing and TDRM analysis. Moreover, this study aimed to focus, for the first time, on women diagnosed with HIV in Israel. Women are a subgroup of patients that was not considered previously a risk group although the biological sex is an important variable determining risk of HIV infection and of subsequent viral pathogenesis as well as of treatment responses (29) .