China is a country with very complex HIV genetic diversity [12]. In the past five years, more than 40 CRFs have been reported in China, including CRF109_0107, CRF110_BC, CRF112_01B, CRF101_01B, and CRF102_0107 [13–16], and many URFs have also been reported by various provinces and cities throughout the country. As the political, economic and cultural center of China, Beijing experiences a continuous flow of people, and the number of PLWHA has been increasing since the first AIDS case was reported in 1985.
This study retrospectively analyzed the relationships among demographic characteristics, genotype diversity, TDR prevalence and the transmission network of newly diagnosed PLWHA in a hospital in Beijing over 9 consecutive years from 2015 to 2023.
First, we observed the changes in the dominant HIV genotype in Beijing. In this study, CRF01_AE dominated from 2015 to 2019, but CRF07_BC surpassed CRF01_AE (36.83% vs. 33.99%) for the first time to become the dominant genotype in 2020; CRF07_BC was slightly lower than CRF01_AE (35.82% vs. 37.80%) in 2021 but then exceeded CRF01_AE again in 2022 and 2023 (2022: 36.99% vs. 35.84%; 2023: 38.79% vs. 34.30%). One study revealed that the number of CRF07_BC infections in PLWHA in Hebei Province in 2021 exceeded that of CRF01_AE for the first time (unpublished data). The fact that Beijing and Hebei share a border also confirms the spatial possibility of a shift in the dominant strain in Beijing.
The present study also revealed that there were more than 20 subtypes and multiple URFs, but CRF01_AE, CRF07_BC and subtype B were still the most common evolutionary branches in Beijing, which is consistent with several studies in the region [17–19]. In addition, CRF01_AE and subtype B showed a gradual downward trend, whereas CRF07_BC showed a rapid upward trend, which was consistent with the overall trend of HIV genotype variation in China [20]. Notably, CRF114_0105, which was first reported by Li Yang et al. [21] in 2021 as a novel CRF originating from Henan Province, was ranked fourth after subtype B in this study. The genome of CRF114_0155, which is the first HIV-1 third-generation CRF and the first CRF55_01B offspring, consists of five segments, three inherited from CRF01_AE cluster 4 and two from CRF55_01B. The emergence of CRF114_0155 reflects the increasing complexity of the HIV-1 genotype. The rapid emergence of this diverse recombinant subtype poses a substantial challenge to the prevention and control of HIV/AIDS in China, making monitoring the molecular epidemiology of HIV-1 critically important.
Drug-resistant variants (DRVs) of HIV-1 may develop spontaneously or may be acquired by transmission in untreated patients. Most of these variants rapidly revert to wild type (WT) without drug selection pressure, but some variants may persist in latent infected cells long after transmission to a new host, limiting the options for ART [22]. Therefore, monitoring drug resistance is critical to clinical care. We next determined the TDR prevalence in Beijing for 9 consecutive years; the overall drug resistance rate was 9.53%, and the rate of NNRTI resistance was 5.38%, which was moderate (5–15%) according to the WHO definition [23]. The overall drug resistance rate and NNRTI and NRTI resistance rates tended to decrease from 2015 to 2017, which was consistent with the findings of Yanze Shi et al. [18]. This result may be related to the ‘Treat for all’ policy that began in 2016. However, from the lowest point in 2017 to 2023, the prevalence of TDR gradually increased, possibly because NNRTIs and NRTIs are the first-line drugs for anti-HIV treatment in China. The prevalence rate of TDR in this study was higher than those reported in Fujian (5.3%) [24], Chongqing (4.08%) [25], Ningbo (6.1%) [26], and Shenzhen (6.02%) [27], which may be related to the high rate of drug resistance in 2015–2016 and to differences in the follow-up and management of PLWHA in different regions and different periods reported in the literature. Compared with that reported from other hospitals in Beijing (from 2013 to 2020) [18], the TDR prevalence rate from 2015 to 2020 was basically consistent.
Among the three kinds of antiretroviral drugs, NNRTIs had the highest prevalence of TDR (Table 2), with the most frequent mutation being V179D (53.61%, 401/748). For NRTIs, the most frequent mutation site was S68, followed by M184, whereas for PIs, the most frequent mutation site was M46. These mutation sites are essentially consistent with those reported by other hospitals in Beijing and Tianjin [18, 28]. V179D/E mutations can lead to a 2–5-fold reduction in susceptibility to NVP, EFV, ETR, and RPV [29], whereas the combination of V179D with other mutations, such as K103R, can increase HIV drug resistance by 16–27 times [30]. S68G often cooccurs with K65R, with a frequency of 21.1–61.7% in various subtypes [31]. Drug sensitivity results revealed that the resistance of S68 to DFC, lamivudine, emtricitabine, tenofovir, and abacavir increased 10–30-fold, and S68 was moderately resistant to stavudine but still sensitive to lopinavir [32]. Mutations in M46 can reduce the sensitivity of PIs (enhanced lopinavir (LPV/r), enhanced atazanavir (ATV/r), and DRV/r)) to HIV [33]. The emergence of these drug resistance sites has greatly reduced the sensitivity of patients to these drugs and severely affected their therapeutic effects. Therefore, real-time monitoring of HIV drug resistance and the related sites, as well as performing drug resistance testing before starting ART, is important for reducing the spread of TDR and has great value for clinicians in the selection of therapeutic drugs.
Finally, we analyzed the important role of genetic networks in identifying individuals at high risk of HIV transmission. We observed demographic differences between the enrolled and unenrolled participants and found that individuals aged 25–34 years and 45–54 years, Han Chinese individuals, infected individuals without TDR, and those with less than college degrees were more likely to be included in the network. Moreover, we analyzed the potential factors of genotype conversion in Beijing through the construction of transmission networks. In the present study, 48 clusters were formed from 222 CRF01_AE sequences, whereas 472 CRF07_BC sequences constituted 8 clusters, suggesting that individuals with the CRF07_BC genotype are more prone to link with other cases and build large clusters, leading to a more extensive spread of CRF07_BC, which is evidenced by the largest growing transmission cluster in our study (Fig. 4D). The large cluster is dominated by CRF07_BC (90.52%, 449/496) and has gradually expanded from 11 cases in 2015 to 496 cases over 9 years with rapid annual growth. Additionally, other genotypes (CRF80_0107, CRF120_0107, CRF121_0107, URF) were also added, indicating that CRF07_BC acts as an active bridge for HIV transmission between different social circles, facilitating the spread of the virus to populations outside the cluster. This feature is consistent with observations in Ningbo [26]. Furthermore, from a demographic perspective, the cases included in this cluster were mainly male (96.57%, 479/496), single (62.70%, 311/496), with a college education or above (58.67%, 291/496) and infected through homosexual transmission (81.45%, 404/496). The proportion of CRF07_BC in individuals aged < 25 years and > 55 years was greater than that of other genotypes (Supplementary materials, Table S2). In addition, during the nine-year period of continuous observation in this study, CRF01_AE clusters occasionally had new cases added, but no clusters with sustained growth were identified. These clusters were mainly small and medium-sized slow-spreading clusters, which is in sharp contrast to the clusters composed of CRF07_BC. This finding likely indicates why CRF07_BC became the dominant strain in Beijing. Therefore, the construction of molecular networks has made the timely detection of key high-risk groups possible. From the perspective of HIV epidemic prevention and control, targeted intervention and supervision of high-risk individuals are urgent measures that can interrupt the rapid spread of HIV [34, 35].
The transmission of TDR is an important factor in assessing cluster risk. Studies have shown that patients with TDR are at a greater risk of being included in the network than are patients without TDR [36]. Among the 76 molecular clusters in this study, 20 (26.31%) contained individuals both with and without TDR or only those with TDR, but no sustained growth was observed in these clusters. In the largest transmission cluster formed by CRF07_BC (Fig. 4D), TDR occasionally appeared but did not constitute the majority of transmission cases. However, given the potential threat of the widespread transmission of TDR, close and continuous monitoring is still needed.
Compared with the size of the networked population, clusters that included multiple transmission routes posed greater communication risk. In the largest CRF07_BC cluster, multiple cases were related to heterosexual transmission (Fig. 4C). Women and men who have sex with both men and women (MSMW) became bridges between cluster members and those outside the cluster. This emergence of multiple and complex transmission routes increases the risk of HIV spread [37], and strict intervention should be implemented for the target individuals.
Limitations of the study
This study had several limitations. First, all the results of this study were obtained from laboratory data and theoretical analysis. The application of targeted interventions in high-risk individuals based only on genetic transmission networks remains insufficient in actual clinical practice. Precise interventions need to be implemented in conjunction with epidemiological investigations for validation. Second, our study population was drawn from a specialized infectious disease hospital in Beijing, and all conclusions are based on nine years of continuous data from this hospital. To avoid potential bias, further studies with larger sample sizes should be conducted.