www.nature.

com/scientificreports

OPEN Effects of HIV‑1 genotype

on baseline CD4+ cell
count and mortality
before and after antiretroviral
therapy
Zhiqiang Cao1, Jianjun Li2, Huanhuan Chen2, Chang Song1, Zhiyong Shen2, Xinjuan Zhou2,
Guanghua Lan2, Qiuying Zhu2, Shujia Liang2, Hui Xing1, Lingjie Liao1, Yi Feng1,
Yiming Shao1 & Yuhua Ruan1*

To assess whether human immunodeficiency virus type 1 (HIV-1) genotype influences baseline CD4+ T
lymphocyte (CD4+) cell count and mortality of patients. The study was conducted from 2014 to 2019
in Guangxi, China, and included 2845 newly diagnosed HIV patients. We used a median regression
model to compare CD4+ cell counts in patients newly diagnosed with different HIV-1 genotypes, and a
Cox regression model to analyze the associations between HIV-1 genotypes and mortality before and
after antiretroviral treatment (ART). In newly diagnosed HIV patients, the baseline CD4+ cell counts
of patients with CRF01_AE were significantly lower than those of patients with CRF07_BC, CRF08_BC,
and other genotypes. Compared with CRF01_AE, patients infected with CRF07_BC (hazard ratio, 0.55;
95% CI 0.36–0.85), CRF08_BC (hazard ratio, 0.67; 95% CI 0.52–0.85), or other genotypes (hazard ratio,
0.52; 95% CI 0.29–0.94) had significantly lower mortality rates before ART. There were no significant
associations between different HIV-1 genotypes and mortality after ART. HIV-1 genotype significantly
influences baseline CD4+ cell count and mortality before ART in newly diagnosed HIV patients. We find
no significant difference in the outcome of death after ART in patients with different HIV-1 genotypes.

Human immunodeficiency virus (HIV) exhibits a broad range of genetic ­heterogeneity1. The HIV epidemic is
mainly caused by HIV-1 group M, which accounts for at least 90% of HIV infections w ­ orldwide2. HIV-1 group
M is further divided into distinct genotypes, namely A, B, C, D, F, G, H, J, and K, as well as hybrids resulting
from recombination between genotypes, known as circulating recombinant forms (CRFs) or unique recombi-
nant forms (URFs)2,3. This genetic diversity is a major obstacle to the development of a vaccine against HIV, as
an effective vaccine must protect against diverse HIV ­genotypes4. Furthermore, HIV genetic diversity may also
affect virus pathogenicity, disease progression, and response to antiretroviral treatment (ART) among infected
­patients1,5. A relatively consistent finding is that compared with genotype A, genotype D is associated with
mortality before ART​6–8. Research on HIV-1 subtypes and ART showed that infection with genotype A results
in increased mortality compared to that with genotype B in Europe and ­Canada9. In China, studies showed that
CRF01_AE is associated with faster disease progression from estimated seroconversion to acquired immune
deficiency syndrome (AIDS) compared with non-CRF01_AE in ART-naïve p ­ atients10,11. These aforementioned
studies revealed that HIV-1 genotype differences contribute to variations in disease progression and ART effi-
ciency. However, studies focused on CRF01_AE featured small participant numbers and short follow-up peri-
ods, which necessitates broader comparisons between one genotype and all other genotypes grouped together.
Furthermore, only a few studies have focused on the CRF07_BC and CRF08_BC genotypes, which are prevalent
in ­China12–14. In addition, the effects of CRF01_AE, CRF07_BC, and CRF08_BC have not been directly assessed

1
State Key Laboratory of Infectious Disease Prevention and Control (SKLID), National Center for AIDS/STD
Control and Prevention (NCAIDS), Chinese Center for Disease Control and Prevention (China CDC), Collaborative
Innovation Center for Diagnosis and Treatment of Infectious Diseases, Beijing, China. 2Guangxi Zhuang
Autonomous Region Center for Disease Prevention and Control, Nanning, China. *email: [email protected]

Scientific Reports | (2020) 10:15875 | https://doi.org/10.1038/s41598-020-72701-4 1

Vol.:(0123456789)
 www.nature.com/scientificreports/

using mortality as a study end point for either ART-naïve participants or those on ART. This question is urgent
with respect to HIV-1 epidemic in China, as transmission accounts for more than 90% of cases, resulting in
infections by CRF07_BC, CRF08_BC, and multilineages of CRF01_AE; this is distinct from the single lineage
epidemic of B genotype prevalent in European ­countries12,13,15,16. These genotypes are also increasing rapidly
among the global proportion of genotypes over t­ ime1,2.
At the end of 2017, more than 70,000 infected patients were living with HIV/AIDS in Guangxi, one of the
most important epicenters of the HIV epidemic in ­China17. Previous studies have shown that the cumulative
mortality of HIV in Guangxi could reach three times that of the national average and up to nine times that of
some developed countries, which is likely associated with the late discovery of HIV ­infection18–22. In recent years,
studies have revealed that the prevalence of HIV-1 CRF01_AE, CRF07_BC, and CRF08_BC is expanding rapidly
in southwestern C­ hina13. Assessing genotypes as risk factors for disease progression and the response to ART is
both urgent and necessary for better understanding the virus characteristics, justifying the early initiation and
efficiency of ART, and reducing mortality and transmission rates. This study was performed to investigate the
effects of HIV-1 genotype on baseline CD4+ cell counts and mortality, before and after ART, among newly diag-
nosed HIV patients. We found that CRF01_AE was significantly associated with a lower baseline CD4+ count
and higher mortality before ART treatment, and that combination ART treatment is equally efficient among
HIV patients regardless of their HIV-1 genotype.

Results
Baseline characteristics of study participants. Of the 6,078 HIV patients who were newly diagnosed
between January 1, 2014 and April 30, 2019, 3237 had blood samples available in storage for HIV-1 genotype
analysis after western blotting (WB) confirmation tests. Some patients (n = 392) were excluded due to nested
polymerase chain reactions (PCR) amplification failure (n = 195), length of the nucleotide sequence being less
than 1000 base pairs (n = 11), mixed bases of nucleotide sequence being more than 5% (n = 9), failure to link
records (n = 16), no CD4+ T lymphocyte (CD4+) cell count measurements (n = 148), or being under the age of
18 years (n = 13). Among 2845 patients who met the eligibility criteria, 2845 HIV patients entered the ART-naïve
cohort (before ART) and 2083 HIV patients were included in the ART cohort (after ART; Fig. 1).
The baseline characteristics of the 2845 patients are shown in Table 1. Most patients were over 50 years of age
(53.1%), male (74.2%), married or cohabiting (60.5%), Han (81.5%), educated up to primary school level or less
(55.3%), employed in farming (78.8%), and infected through heterosexual contact (94.2%). The median time
from HIV-positive diagnosis to baseline CD4+ cell count measurements was 7 days [interquartile range (IQR),
2–23 days; Table 1]. A phylogenetic genotype tree is shown in Fig. 2. The baseline genotyping data of the 2845
study participants revealed that the predominant genotype was CRF01_AE for 1687 patients (59.3%; 1221 with
CRF01_AE Cluster 2, 331 with CRF01_AE Cluster 1, and 135 with other clusters), followed by CRF08_BC for
728 patients (25.6%), CRF07_BC for 254 patients (8.9%), and other genotypes for 176 patients (6.2%; 59 with
various other CRFs, 68 with URFs, 37 with CRF55_01B, and 12 with B/B’).

Effects of HIV‑1 genotype on baseline CD4+ cell counts among newly diagnosed HIV
patients. Table 2 presents the unadjusted and adjusted effects of HIV-1 genotype on baseline CD4+ cell
counts in newly diagnosed HIV patients. Among 2845 patients, the median (IQR) baseline CD4+ cell count was
111 cells/μl (31–279), 250 cells/μl (156–377), 205 cells/μl (108–327), and 198 cells/μl (89–355) for CRF01_AE,
CRF07_BC, CRF08_BC, and other genotypes, respectively. After adjusting for the potential effects of other vari-
ables in a multivariable median regression model, the coefficient of baseline CD4+ cell count was 113 (95% CI
86–140) in CRF07_BC, 80 (95% CI 65–95) in CRF08_BC, and 72 (95% CI 37–107) in the other genotypes group,
which was significantly higher than that in the CRF01_AE group.

Effects of HIV‑1 genotype on mortality before ART among newly diagnosed HIV
patients. Table 3 presents the unadjusted and adjusted effects of HIV-1 genotype on mortality among newly
diagnosed HIV patients in the ART-naïve cohort between 2014 and 2019. There were 451 deaths in this group
and a total of 1614.2 person-years were recorded during follow-up. The overall mortality was 31.8 deaths (95%
CI 29.0–34.9) per 100 person-years among the study patients. The genotype-associated mortality was 42.6 (95%
CI 38.1–47.5), 16.5 (95% CI 10.6–24.5), 21.9 (95% CI 17.6–26.9), and 12.6 (95% CI 6.5–22.1) deaths per 100
person-years for patients with CRF01_AE, CRF07_BC, CRF08_BC, and other genotypes, respectively. Before
ART, the hazard ratios for the association between the genotype and mortality were almost identical in both
univariate and multivariate models. In multivariate analyses, compared to that in HIV patients infected with
CRF01_AE, patients infected with CRF07_BC (hazard ratio, 0.55; 95% CI 0.36–0.85), CRF08_BC (hazard ratio,
0.67; 95% CI 0.52–0.85), and other genotypes (hazard ratio, 0.52; 95% CI 0.29–0.94) had a significantly lower
mortality rate before ART, respectively.

Effects of HIV‑1 genotype on mortality before ART among newly diagnosed HIV patients with
baseline CD4+ cell counts < 350 cells/μl. Based on our data showing that HIV patients infected with
CRF01_AE had lower baseline CD4+ cell counts (cells/μl) than those with other HIV-1 genotypes, we speculated
that this might lead to a higher death rate before ART. Table 4 presents the subgroup analysis of unadjusted and
adjusted effects of HIV-1 genotype on mortality among newly diagnosed HIV patients with a baseline CD4+ cell
count of < 350 cells/μl in the ART-naïve cohort. A total of 424 deaths occurred in this group and 934.9 person-
years were recorded during follow-up. The overall mortality was 45.3 deaths (95% CI 41.1–49.9) per 100 person-
years among the study patients. The genotype-associated mortality was 57.5 (95% CI 51.1–64.4), 26.4 (95%
CI 16.2–40.3), 32.0 (95% CI 26.6–40.7), and 20.9 (95% CI 10.9–36.2) deaths per 100 person-years for patients

Scientific Reports | (2020) 10:15875 | https://doi.org/10.1038/s41598-020-72701-4 2

Vol:.(1234567890)
www.nature.com/scientificreports/

6078 HIV patients diagnosed by WB between

January 1, 2014 and April 30, 2019

3237 HIV patient stored blood samples for

HIV-1 genotype analysis remaining after WB

392 excluded
195 PCR amplification failed
11 length of nucleotide sequence < 1000 bp
9 mixed bases of nucleotide sequence > 5%
16 failure to link records
148 with no CD4+ measurements
13 < 18 years of age

2845 HIV patients included in cohort

762 never receiving ART excluded

2845 HIV patients included in 2083 HIV patients included in

ART-naïve cohort (before ART) ART cohort (after ART)

Outcome Outcome
451 died 163 died
90 lost to follow-up 210 lost to follow-up
2083 started ART 21 stopped ART
221 survived 1689 survived

Figure 1.  Flow diagram of study. ART​antiretroviral therapy, WB western blotting. We performed this
observational cohort study using data from the HIV/AIDS Comprehensive Response Information Management
System (CRIMS) in Guangxi, China. Local health workers completed standardized reporting forms recording
baseline information when HIV patients were newly diagnosed, and follow-up visits then conducted.

with CRF01_AE, CRF07_BC, CRF08_BC, and other genotypes, respectively. The hazard ratios for the asso-
ciation between the genotype and mortality were almost identical in both univariate and multivariate models.
The resulting multivariate modeling showed that compared to that for HIV patients infected with CRF01_AE,
patients infected with CRF07_BC (hazard ratio, 0.49; 95% CI 0.31–0.76), CRF08_BC (hazard ratio, 0.60; 95% CI
0.47–0.77), and other genotypes (hazard ratio, 0.46; 95% CI 0.26–0.82) had a significantly lower mortality rate
before ART, respectively (Table 4).

Effects of HIV‑1 genotype on mortality after ART among newly diagnosed HIV patients. Table 5
presents the unadjusted and adjusted effects of HIV-1 genotype on mortality among newly diagnosed HIV
patients in the ART cohort between 2014 and 2019. In this cohort, 163 deaths occurred and a total of 4301.8
person-years were recorded during follow-up. The overall mortality was 3.8 deaths (95% CI 3.2–4.4) per 100
person-years among the study patients. The genotype-associated mortality was 4.1 (95% CI 3.4–4.9), 2.2 (95% CI
0.9–4.2), 3.9 (95% CI 2.8–5.4), and 2.1 (95% CI 0.7–4.8) deaths per 100 person-years for patients with CRF01_
AE, CRF07_BC, CRF08_BC, and other genotypes, respectively. The results from both univariate and multivari-
ate analyses showed that no genotypes were significantly associated with death in this group (Table 5).

Scientific Reports | (2020) 10:15875 | https://doi.org/10.1038/s41598-020-72701-4 3

Vol.:(0123456789)
 www.nature.com/scientificreports/

HIV-1 genotype
Total CRF01_AE CRF07_BC CRF08_BC Others#
Variable N (%) N (%) N (%) N (%) N (%) P
Total 2845 (100.0) 1687 (59.3) 254 (8.9) 728 (25.6) 176 (6.2)
Age at diagnosis, years
18–29 279 (9.8) 143 (8.5) 49 (19.3) 57 (7.8) 30 (17.0) < 0.001
30–49 1054 (37.1) 593 (35.1) 62 (24.4) 318 (43.7) 81 (46.1)
≥ 50 1512 (53.1) 951 (56.4) 143 (56.3) 353 (48.5) 65 (36.9)
Sex 0.001
Male 2112 (74.2) 1257 (74.5) 200 (78.7) 510 (70.1) 145 (82.4)
Female 733 (25.8) 430 (25.5) 54 (21.3) 218 (29.9) 31 (17.6)
Marital status < 0.001
Married or cohabiting 1722 (60.5) 1038 (61.5) 144 (56.7) 447 (61.4) 93 (52.8)
Single 576 (20.2) 319 (18.9) 67 (26.4) 132 (18.1) 58 (33.0)
Divorced or widowed 547 (19.2) 330 (19.6) 43 (16.9) 149 (20.5) 25 (14.2)
Ethnicity < 0.001
Han 2320 (81.5) 1360 (80.6) 208 (81.8) 626 (86.0) 126 (71.6)
Zhuang 250 (8.8) 172 (10.2) 23 (9.1) 39 (5.4) 16 (9.1)
Other 275 (9.7) 155 (9.2) 23 (9.1) 63 (8.6) 34 (19.3)
Education level 0.003
Primary school or less 1572 (55.3) 949 (56.3) 123 (48.4) 420 (57.7) 80 (45.5)
Junior middle school or more 1273 (44.7) 738 (43.7) 131 (51.6) 308 (42.3) 96 (54.5)
Occupation < 0.001
Farm 2243 (78.8) 1336 (79.2) 163 (64.2) 614 (84.3) 130 (73.9)
None-farm 602 (21.2) 351 (20.8) 91 (35.8) 114 (15.7) 46 (26.1)
Transmission route < 0.001
Heterosexual contact 2680 (94.2) 1630 (96.6) 221 (87.0) 679 (93.3) 150 (85.2)
Male-to-male sexual contact 64 (2.2) 21 (1.2) 28 (11.0) 0 (0.0) 15 (8.5)
Injecting drug use 101 (3.6) 36 (2.1) 5 (2.0) 49 (6.7) 11 (6.3)
Year of HIV infection diagnosis < 0.001
2014–2015 853 (30.0) 571 (33.8) 58 (22.8) 193 (26.5) 31 (17.6)
2016–2017 1208 (42.5) 716 (42.4) 120 (47.2) 298 (40.9) 74 (42.0)
2018–2019 784 (27.5) 400 (23.7) 76 (29.9) 237 (32.6) 71 (40.3)
Time from HIV diagnosis to baseline CD4+ cell count
7 (2–23) 8 (2–25) 6 (2–18) 6 (2–19) 8 (3–23) < 0.001
measurements, days

Table 1.  Baseline characteristics of study participants. Data are presented as No. (%) or median (interquartile
range). χ2 test for categorical data and non-parametric Wilcoxon test for continuous data. # HIV-1 B/B’,
CRF55_01B, CRFs and URFs.

Discussion
In the ART-naïve cohort, there were significant associations between HIV patients with CRF01_AE and lower
baseline CD4+ cell count at diagnosis, as well as higher mortality, compared with CRF07_BC, CRF08_BC, and
other genotypes. A previous study showed that the baseline CD4+ cell count in CRF01_AE patients was lower
than that in non-CRF01_AE patients for men who have sex with ­men23. Our study further demonstrates that
CRF01_AE patients have a lower CD4+ cell count than those with CRF07_BC, CRF08_BC, or other genotypes
at initial diagnosis in a large population with different transmission routes. Previous studies have suggested that
CRF01_AE is associated with faster AIDS progression (defined as CD4+ T-cell counts decreasing to < 200 cells/
μl or < 350 cells/μl) compared to that with non-CRF01_AE10,11. We demonstrate for the first time that patients
with CRF01_AE exhibit a significantly higher mortality compared than those with CRF07_BC, CRF08_BC,
and other genotypes, in a large observational cohort study. Some studies have described that a lower baseline
CD4+ cell count is associated with a risk of higher mortality compared to a higher baseline CD4+ cell count in
HIV ­patients24–27. The CD4+ cell count and immune status are intermediate variables between genotypes and
deaths from the perspective of epidemiological causality. To further avoid bias caused by different genotypes
with heterogeneous baseline CD4+ cell counts, we performed sensitivityanalysis using a multivariate Cox model
grouped by CD4+ cell count and obtained consistent results. It has been suggested that the CRF01_AE genotype
is an independent risk factor for mortality in ART-naïve patients, compared with the CRF07_BC, CRF08_BC,
and other genotypes. The entry of HIV into immune cells is highly dependent on the interaction of the viral
envelope (R5 tropism and/or X4 tropism) with its target co-receptor (CCR5 and/or CXCR4)28, and an accelerated
CD4+ cell count decline appears to be driven by an elevated proportion of CD4+ cells expressing C ­ XCR429–31.

Scientific Reports | (2020) 10:15875 | https://doi.org/10.1038/s41598-020-72701-4 4

Vol:.(1234567890)
www.nature.com/scientificreports/

Study strain
Refrence strain
g2a_AE.VN.97VNAG201
CRF01_AE
Cluster 2
Cluster 1
CRF55_01B
CRF08_BC
CRF07_BC g2_AE.CN.FJ070013

B/B’

g1_AE.CN.05GX012

55_01B.CN.HNCS102056

08_BC.97CNGX

07_BC.CN.HEB.X.1114
07_BC.CN.XJ..97CN001

B.MM.mSTD101.AB097870
B.CN.SM.HQ215554
0.03

Figure 2.  Phylogenetic tree of HIV-1 Pol. Shimodaira–Hasegawa test was used to alternate the topologies of the
whole sequences. Branches, representing their respective sequences, forming node with high confidence (> 90%)
was defined as the same type of genotype. The scale length (range: 0–1) represents the difference between
sequences (range: 0–100%).

Scientific Reports | (2020) 10:15875 | https://doi.org/10.1038/s41598-020-72701-4 5

Vol.:(0123456789)
 www.nature.com/scientificreports/

Unadjusted Adjusted
HIV-1 Genotype Patients, no. (%) CD4+ median (IQR) Coefficient 95% CI P Coefficient 95% CI P
Total 2845 (100.0) 160 (49–310)
CRF01_AE 1687 (59.3) 111 (31–279) 0 0
CRF07_BC 254 (8.9) 250 (156–377) 139 (117–161) < 0.001 113 (86–140) < 0.001
CRF08_BC 728 (25.6) 205 (108–327) 94 (73–115) < 0.001 80 (65–95) < 0.001
Others 176 (6.2) 198 (89–355) 87 (46–128) < 0.001 72 (37–107) < 0.001

Table 2.  Effects of HIV-1 genotype on baseline CD4+ cell counts (cells/μl) among newly diagnosed HIV
patients. Adjusted for age at diagnosis, sex, marital status, ethnicity, education level, occupation, transmission
route, and year of HIV infection diagnosis.

Deaths/100
person-years
HIV-1 genotype Patients, no Death, no Person-years (95% CI) HR (95% CI) P AHR (95% CI) P
Total 2845 451 1614.2 31.8 (29.0–34.9)
CRF01_AE 1687 324 760.4 42.6 (38.1–47.5) 1.00 1.00
CRF07_BC 254 24 145.7 16.5 (10.6–24.5) 0.42 (0.28–0.63) < 0.001 0.55 (0.36–0.85) 0.006
CRF08_BC 728 91 415.2 21.9 (17.6–26.9) 0.56 (0.44–0.71) < 0.001 0.67 (0.52–0.85) 0.001
Others 176 12 95.0 12.6 (6.5–22.1) 0.33 (0.18–0.58) < 0.001 0.52 (0.29–0.94) 0.030

Table 3.  Effects of HIV-1 genotype on mortality before antiretroviral therapy among newly diagnosed HIV
patients. HR (Hazard ratios) were calculated by means of both univariate and multivariate Cox regression
analysis, adjusted for age at diagnosis, sex, marital status, ethnicity, education level, occupation, transmission
route, year of HIV infection diagnosis, and baseline CD4+ cell count per 100 increment.

Deaths/100
person-years
HIV-1 genotype Patients, no Death, no Person-years (95% CI) HR (95% CI) P AHR (95% CI) P
Total 2284 424 934.9 45.3 (41.1–49.9)
CRF01_AE 1409 303 529.0 57.5 (51.1–64.4) 1.00
CRF07_BC 176 21 78.9 26.4 (16.2–40.3) 0.49 (0.32–0.77) 0.002 0.49 (0.31–0.76) 0.002
CRF08_BC 571 88 269.6 32.0 (26.6–40.7) 0.62 (0.49–0.79) < 0.001 0.60 (0.47–0.77) < 0.001
Others 129 12 57.4 20.9 (10.9–36.2) 0.40 (0.22–0.71) 0.002 0.46 (0.26–0.82) 0.009

Table 4.  Effects of HIV-1 genotype on mortality before antiretroviral therapy among newly diagnosed HIV
patients with baseline CD4+ cell counts < 350 cells/μl. HR (Hazard ratios) were calculated by means of both
univariate and multivariate Cox regression analysis, adjusted for age at diagnosis, sex, marital status, ethnicity,
education level, occupation, transmission route, year of HIV infection diagnosis, and baseline CD4+ cell count
per 100 increment.

Deaths/100
person-years
HIV-1 Genotype Patients, No Death, No Person- years (95%CI) HR (95% CI) P AHR (95% CI) P
Total 2083 163 4301.8 3.8 (3.2–4.4)
CRF01_AE 1,217 111 2699.4 4.1 (3.4–4.9) 1.00 1.00
CRF07_BC 192 8 367.2 2.2 (0.9–4.2) 0.54 (0.26–1.10) 0.088 0.78 (0.37–1.62) 0.501
CRF08_BC 535 39 993.7 3.9 (2.8–5.4) 0.96 (0.67–1.39) 0.833 1.27 (0.86–1.87) 0.223
Others 139 5 241.5 2.1 (0.7–4.8) 0.51 (0.21–1.26) 0.144 0.95 (0.38–2.39) 0.909

Table 5.  Effects of HIV-1 genotype on mortality after antiretroviral therapy (ART) among newly diagnosed
HIV patients. HR (Hazard ratios) were calculated by means of both univariate and multivariate Cox regression
analysis, adjusted for age at diagnosis, sex, marital status, ethnicity, education level, occupation, transmission
route, year of HIV infection diagnosis, CD4+ cell count per 100 increment before ART, and antiretroviral
regimens (grouped by zidovudine-based regimen, tenofovir-based regimen, and other regimens).

Scientific Reports | (2020) 10:15875 | https://doi.org/10.1038/s41598-020-72701-4 6

Vol:.(1234567890)
www.nature.com/scientificreports/

Researchers have shown that CRF01_AE has a higher degree of X4 tropism than non-CRF01_AE in patients
infected via sexual ­transmission10.
In the ‘after ART’ cohort, our analyses revealed no significant difference in mortality among HIV patients with
different genotypes, in either univariate or multivariate Cox models. In this study, the 6-year follow-up mortality
of HIV patients after ART was an average of 3.8 deaths per 100 person-years, which is slightly higher than the
mortality rates of some developed countries, such as England and C ­ anada20,21. Previous studies have shown that
there is no significant difference in immune recovery between CRF01_AE and B strains in a combined ART
­cohort32. Recently, researchers found that compared to that with CRF07_BC, CRF01_AE is associated with a
lower CD4+ cell count recovery and a slower rate of immune recovery in combined ART​33. Our study is the first
to show that there is no difference in the therapeutic efficacy of combined ART among patients with CRF01_AE,
CRF07_BC, CRF08_BC, and other genotypes, using death as the endpoint. Previous studies have shown that
AZT exhibits a higher efficacy against viruses with R5 rather than X4 tropism. This might be explained by the
fact that CCR5-expressing activated memory T cells demonstate a high efficiency in the activation of reverse
transcriptase ­inhibitors34–36. Some nucleoside reverse-transcriptase inhibitors, like 3TC, are activated equally
in both CCR5 and CXCR4 of target cells, and drugs do not require activation, such as the protease inhibitor
ritonavir, show equal activity against both CCR5- and CXCR4-specific v­ ariants28,36,37.
This study showed that CRF01_AE, CRF07_BC, and CRF08_BC were the dominant HIV genotypes in
Guangxi from 2014 to 2019. CRF01_AE comprised the largest proportion of the endemic strains and was com-
posed mainly of CRF01_AE cluster 1 and cluster 2. It is worth noting that prostitutes from Vietnam have been
active in the southern border area of Guangxi in recent years. CRF01_AE cluster 2, which accounts for the biggest
proportion of the strains comprising the local epidemic in southern Guangxi, shows high genetic homology with
the Vietnamese reference strain, according to a previous s­ tudy38. CRF01_AE cluster 1, which is mainly prevalent
in northern Guangxi, was reported to have originated in Th ­ ailand16,39. Our phylogenetic tree of HIV strains from
the Guangxi epidemic show that these strains have formed many short branches in their evolutionary lineage,
suggesting that the HIV epidemic in this area spread rapidly over a short period.
Several limitations of this study should be noted. First, censoring of information caused by loss to follow-
up and ART initiation might lead to informative bias regarding the mortality rates of the ART-naïve cohort,
particularly if the censoring was not random. Nevertheless, we have made an effort to measure (and include in
the model) any covariates that are likely to affect the rate of censoring. Furthermore, we did not observe any
differences between genotypes regarding rates of loss to follow-up and ART initiation in the ART-naïve cohort.
Second, in this study, the CRF01_AE strains cluster 1 and cluster 2 comprised the bulk of the CRF01_AE infec-
tions in Guangxi, which does not cover all clusters of CRF01_AE found in China. It should be noted that we did
not find any difference in mortality between CRF01_AE clusters 1 and 2 (data not shown).
However, these limitations do not invalidate our conclusions. This is the first large cohort study in a real-
world setting to evaluate the effects of CRF01_AE, CRF07_BC, CRF08_BC, and other genotypes on baseline
CD4+ cell counts and mortality before and after ART among newly diagnosed HIV patients. This study revealed
that there is a high risk of AIDS mortality for HIV patients due to the rapid progression of CRF01_AE in ART-
naïve patients, and showed that combination ART is equally efficient among HIV patients regardless of HIV-1
genotype. Our findings highlight the significance of early detection and timely treatment to reduce the risk of
AIDS-related mortality.

Materials and methods

Study population and design. We performed an observational cohort study using data from the HIV/
AIDS Comprehensive Response Information Management System (CRIMS), a Web-based real-time informa-
tion collection and maintenance database for the national HIV epidemic in China, which has been described
­elsewhere40–42. Local health workers completed standardized reporting forms recording baseline information
when HIV patients were newly diagnosed, and follow-up visits were conducted every 3 months before ART. For
patients starting ART, follow-up visits occurred at 0.5, 1, 2, and 3 months, and then every 3 months thereafter.
This study focused on patients from Guangxi, rural Southwest China. Data were collected from January 1, 2014
to April 30, 2019. The eligibility criteria were HIV-1 patients newly diagnosed between January 1, 2014 and April
30, 2019, age ≥ 18 years, baseline CD4+ cell count measurements, and willingness to participate based on writ-
ten informed consent when entering CRIMS. Patients with unqualified HIV-1 sequences or the inability to link
records were excluded.
Typically, baseline CD4+ cell counts would be measured immediately after patients are newly diagnosed as
HIV positive. Since 2014, the national criteria for starting ART are (1) CD4+ cell count < 500 cells/μl, (2) AIDS
progression to III/IV stage, or (3) willingness to receive ART, regardless of CD4+ cell counts or AIDS progres-
sion. In 2016, these guidelines were changed to offer free ART to all HIV-infected patients, regardless of CD4+
cell counts, with the provision of informed consent. According to the WHO guidelines, the current first-line
combined ART regimens in China include tenofovir (TDF) or zidovudine (AZT) with lamivudine (3TC) and
efavirenz or nevirapine. The second-line ART regimens include TDF or AZT with 3TC and lopinavir-ritonavir43.

Procedures. Our study assessed three outcomes as follows: baseline CD4+ cell count at the time of diagno-
sis, mortality before ART (ART-naïve status), and mortality after ART (ART status).
The baseline characteristics and CD4+ cell counts of newly diagnosed patients in Guangxi were extracted from
the CRIMS database. Patients were successively stratified by treatment status into ART-naïve and ART cohorts,
according to ART initiation time. The ART-naïve cohort included HIV patients during the period spanning the
date of diagnosis to the date of starting ART, or for patients who did not start ART, to the end of April 30, 2019.
The ART cohort included HIV patients from the date of starting ART to April 30, 2019. To assess mortality

Scientific Reports | (2020) 10:15875 | https://doi.org/10.1038/s41598-020-72701-4 7

Vol.:(0123456789)
 www.nature.com/scientificreports/

before and after ART as a life-long process, patients who received ART after being newly diagnosed could be
initially placed in the ART-naïve cohort and then moved to the ART cohort, as these groups were separated by
ART initiation time.
The variables of the baseline characteristics included age at diagnosis, sex, marital status, ethnicity, education
level, occupation, transmission route, year of HIV infection diagnosis, and time from diagnosis to baseline CD4+
cell count measurements. At follow-up, the following characteristics were assessed: duration of ART-naivety, date
of ART initiation, initial ART regimen, duration of ART, cessation of ART, and survival status.
To analyze the effects of HIV-1 genotype on mortality before or after ART among newly diagnosed HIV
patients, we used the time to death as a study endpoint. For ART-naïve patients, time zero was defined as the
date of HIV diagnosis. Data were censored on the date of ART initiation or on April 30, 2019, the study endpoint.
For ART patients, time zero was defined as the date of starting ART and data were censored on April 30, 2019.

Laboratory methods and genotype analysis. HIV-1 RNA was extracted from 200 µl of stored blood
samples, remaining after western blot (WB) confirmation tests, using the QIAamp Viral RNA Mini kit (Qiagen,
Hilden, Germany) according to the manufacturer’s protocol. This HIV-1 RNA pol region encodes regions of
the reverse transcriptase and protease genes (HBX2: 2253–3553), which were amplified by nested polymerase
chain reaction (PCR) with commercial primers according to previously published m ­ ethods44,45. The PCR prod-
ucts were purified using a QIAquick Gel Extraction Kit (Qiagen, Hilden, Germany) and sequenced by Sanger
sequencing on an ABI3730xl sequencer (Life Technologies, Foster City, CA, USA) with the BigDye Terminator
v3.1 kit (Life Technologies, Foster City, CA, USA). Unqualified sequences (length of nucleotide sequence < 1000
base pairs, mixed bases of nucleotide sequence > 5%) were excluded as per laboratory regulations.
The Fasttree 2.146 software program was used to estimate an approximate maximum likelihood phyloge-
netic tree for the pol sequences, together with reference sequences, using a general time reversible model with
g-distributed (G4) among-site rate heterogeneity. The Fasttree program made use of the Shimodaira-Hasegawa
test to alternate topologies of the whole sequences and calculate the confidence of each node in phylogenetic
trees. The final maximum likelihood tree was visualized using the FigTree software program, v1.4.2 (https​://tree.
bio.ed.ac.uk/softw​are/figtr​ee/).

Statistical analysis. We analyzed differences between the various genotypes using χ2 tests (for categorical
data) and non-parametric Wilcoxon tests (for continuous data). We estimated the median baseline CD4+ cell
count by genotype among newly diagnosed HIV patients using univariate and multivariable median regression.
Mortality was calculated based on Poisson distributions, in units of mortality per 100 person-years of follow-up.
We performed both univariate and multivariate Cox proportional hazard models to evaluate the effects of HIV-1
genotype on mortality before and after ART among newly diagnosed HIV patients, respectively. To control for
potential bias and avoid the influences of other variables, the baseline characteristics of age at diagnosis, sex,
marital status, ethnicity, education level, occupation, transmission route, year of HIV infection diagnosis, and
baseline CD4+ cell count were included as control variables in the multivariate median regression model and
multivariate Cox proportional hazard ­model23,47,48.
A two-sided p-value of 0.05 or less was regarded as statistically significant. All analyses were performed using
SAS 9.4 (SAS Institute, Inc., Cary, NC, USA).

Ethics statement. This study was reviewed and approved by the institutional review board of the National
Center for AIDS/STD Control and Prevention (NCAIDS), China CDC. The requirement for informed consent
to participate in the study was considered unnecessary, as all patients provided informed consent when entering
CRIMS at the time of diagnosis.

Data availability
Guangxi Center for Disease Control and Prevention permitted researchers to use the database in this study.

Received: 16 June 2020; Accepted: 4 September 2020

References
1. Taylor, B. S., Sobieszczyk, M. E., McCutchan, F. E. & Hammer, S. M. The challenge of HIV-1 subtype diversity. N. Engl. J. Med. 358,
1590–1602. https​://doi.org/10.1056/NEJMr​a0706​737 (2008).
2. Hemelaar, J. et al. Global and regional molecular epidemiology of HIV-1, 1990–2015: A systematic review, global survey, and trend
analysis. Lancet Infect. Dis. 19, 143–155. https​://doi.org/10.1016/s1473​-3099(18)30647​-9 (2019).
3. Robertson, D. L. et al. HIV-1 nomenclature proposal. Science 288, 55–56. https​://doi.org/10.1126/scien​ce.288.5463.55d (2000).
4. Girard, M. et al. Failure of a human immunodeficiency virus type 1 (HIV-1) subtype B-derived vaccine to prevent infection of
chimpanzees by an HIV-1 subtype E strain. J. Virol. 70, 8229–8233 (1996).
5. Group, T. S. Global epidemiology of drug resistance after failure of WHO recommended first-line regimens for adult HIV-1 infec-
tion: A multicentre retrospective cohort study. Lancet Infect. Dis. 16, 565–575. https​://doi.org/10.1016/s1473​-3099(15)00536​-8
(2016).
6. Kiwanuka, N. et al. Effect of human immunodeficiency virus Type 1 (HIV-1) subtype on disease progression in persons from
Rakai, Uganda, with incident HIV-1 infection. J. Infect. Dis. 197, 707–713. https​://doi.org/10.1086/52741​6 (2008).
7. Pontiano, K. et al. Effect of human immunodeficiency virus (HIV) Type 1 envelope subtypes A and D on disease progression in a
large cohort of HIV-1—positive persons in Uganda. J. Infect. Dis. 9, 2 (2002).
8. Vasan, A. et al. Different rates of disease progression of HIV type 1 infection in Tanzania based on infecting subtype. Clin. Infect.
Dis. 42, 843–852. https​://doi.org/10.1086/49995​2 (2006).

Scientific Reports | (2020) 10:15875 | https://doi.org/10.1038/s41598-020-72701-4 8

Vol:.(1234567890)
www.nature.com/scientificreports/

9. Antiretroviral Therapy Cohort C, Canadian Observational Cohort C, Study UK & Collaboration of Observational HI. Mortality
of treated HIV-1 positive individuals according to viral subtype in Europe and Canada collaborative cohort analysis. AIDS 30,
503–513. https​://doi.org/10.1097/QAD.00000​00000​00094​1 (2016).
10. Li, Y. et al. CRF01_AE subtype is associated with X4 tropism and fast HIV progression in Chinese patients infected through sexual
transmission. AIDS 28, 521–530. https​://doi.org/10.1097/qad.00000​00000​00012​5 (2014).
11. Chu, M. et al. HIV-1 CRF01_AE strain is associated with faster HIV/AIDS progression in Jiangsu Province, China. Sci. Rep. 7,
1570. https​://doi.org/10.1038/s4159​8-017-01858​-2 (2017).
12. Su, Y., Liu, H., Wu, J., Zhu, L. & Wang, N. Distribution of HIV-1 genotypes in China: A systematic review. Zhonghua Liu Xing Bing
Xue Za Zhi 35, 1164–1168 (2014).
13. Li, L. et al. Different distribution of HIV-1 subtype and drug resistance were found among treatment naive individuals in Henan,
Guangxi, and Yunnan province of China. PLoS ONE 8, e75777. https​://doi.org/10.1371/journ​al.pone.00757​77 (2013).
14. Li, J. et al. HIV-1 transmissions among recently infected individuals in Southwest China are predominantly derived from circulat-
ing local strains. Sci. Rep. 8, 12831–12831. https​://doi.org/10.1038/s4159​8-018-29201​-3 (2018).
15. Wang, X. et al. Phylodynamics of major CRF01_AE epidemic clusters circulating in mainland of China. Sci. Rep. 7, 6330. https​://
doi.org/10.1038/s4159​8-017-06573​-6 (2017).
16. Feng, Y. et al. The rapidly expanding CRF01_AE epidemic in China is driven by multiple lineages of HIV-1 viruses introduced in
the 1990s. AIDS 27, 1793–1802. https​://doi.org/10.1097/QAD.0b013​e3283​60db2​d (2013).
17. Xiuling, W. et al. Current situation of follow-up management of the living HIV/-AIDS in Guangxi in 2017. Chin. J. AIDS STD 25,
368–370. https​://doi.org/10.13419​/j.cnki.aids.2019.04.11 (2019).
18. Lin, Z. et al. Studying on the status of mortality of HIV/AIDS patients in Qinzhou city and its influencing factors. Chin. Health
Serv. Manag. 36, 149–152 (2019).
19. Li, M. et al. Mortality among people living with HIV and AIDS in China: Implications for enhancing linkage. Sci. Rep. 6, 28005.
https​://doi.org/10.1038/srep2​8005 (2016).
20. Croxford, S. et al. Mortality and causes of death in people diagnosed with HIV in the era of highly active antiretroviral therapy
compared with the general population: An analysis of a national observational cohort. Lancet Public Health 2, e35–e46. https​://
doi.org/10.1016/s2468​-2667(16)30020​-2 (2017).
21. Eyawo, O. et al. Changes in mortality rates and causes of death in a population-based cohort of persons living with and without
HIV from 1996 to 2012. BMC Infect. Dis. 17, 174. https​://doi.org/10.1186/s1287​9-017-2254-7 (2017).
22. Garriga, C. et al. Mortality, causes of death and associated factors relate to a large HIV population-based cohort. PLoS ONE 10,
e0145701. https​://doi.org/10.1371/journ​al.pone.01457​01 (2015).
23. Li, X. et al. HIV-1 genetic diversity and its impact on baseline CD4+T cells and viral loads among recently infected men who have
sex with men in Shanghai, China. PLoS ONE 10, e0129559. https​://doi.org/10.1371/journ​al.pone.01295​59 (2015).
24. Tang, Z. et al. Effects of high CD4 cell counts on death and attrition among HIV patients receiving antiretroviral treatment: An
observational cohort study. Sci. Rep. 7, 3129. https​://doi.org/10.1038/s4159​8-017-03384​-7 (2017).
25. Nsanzimana, S. et al. Effect of baseline CD4 cell count at linkage to HIV care and at initiation of antiretroviral therapy on mortal-
ity in HIV-positive adult patients in Rwanda: A nationwide cohort study. Lancet HIV 2, e376-384. https​://doi.org/10.1016/s2352​
-3018(15)00112​-5 (2015).
26. May, M. T. et al. Mortality according to CD4 count at start of combination antiretroviral therapy among HIV-infected patients fol-
lowed for up to 15 years after start of treatment: Collaborative cohort study. Clin. Infect. Dis. 62, 1571–1577. https:​ //doi.org/10.1093/
cid/ciw18​3 (2016).
27. Song, A. et al. From CD4-based initiation to treating all HIV-infected adults immediately: An evidence-based meta-analysis. Front.
Immunol. 9, 212. https​://doi.org/10.3389/fimmu​.2018.00212​ (2018).
28. Schuitemaker, H., vant Wout, A. B. & Lusso, P. Clinical significance of HIV-1 coreceptor usage. J. Transl. Med. 9, 5. https​://doi.
org/10.1186/1479-5876-9-s1-s5 (2011).
29. Berkowitz, R. D., Beckerman, K. P., Schall, T. J. & McCune, J. M. CXCR4 and CCR5 expression delineates targets for HIV-1 disrup-
tion of T cell differentiation. J. Immunol. 161, 3702–3710 (1998).
30. Bleul, C. C., Wu, L., Hoxie, J. A., Springer, T. A. & Mackay, C. R. The HIV coreceptors CXCR4 and CCR5 are differentially expressed
and regulated on human T lymphocytes. Proc. Natl. Acad. Sci. USA. 94, 1925–1930. https:​ //doi.org/10.1073/pnas.94.5.1925 (1997).
31. de Roda Husman, A. M., Blaak, H., Brouwer, M. & Schuitemaker, H. CC chemokine receptor 5 cell-surface expression in relation
to CC chemokine receptor 5 genotype and the clinical course of HIV-1 infection. J. Immunol. 163, 4597–4603 (1999).
32. Chow, W. Z. et al. Impact of HIV-1 subtype on the time to CD4+ T-cell recovery in combination antiretroviral therapy (cART)-
experienced patients. PLoS ONE 10, e0137281. https​://doi.org/10.1371/journ​al.pone.01372​81 (2015).
33. Ge, Z. et al. CRF01_AE and CRF01_AE cluster 4 are associated with poor immune recovery in Chinese patients under cART. Clin.
Infect. Dis. https​://doi.org/10.1093/cid/ciaa3​80 (2020).
34. Koot, M. et al. Viral phenotype and T cell reactivity in human immunodeficiency virus type 1-infected asymptomatic men treated
with zidovudine. J. Infect. Dis. 168, 733–736. https​://doi.org/10.1093/infdi​s/168.3.733 (1993).
35. vant Wout, A. B. et al. Changes in cellular virus load and zidovudine resistance of syncytium-inducing and non-syncytium-inducing
human immunodeficiency virus populations under zidovudine pressure: a clonal analysis. J. Infect. Dis. 174, 845–849. https​://doi.
org/10.1093/infdi​s/174.4.845 (1996).
36. van Wout, A. B. et al. Selective inhibition of syncytium-inducing and nonsyncytium-inducing HIV-1 variants in individuals
receiving didanosine or zidovudine respectively. J. Clin. Invest. 100, 2325–2332. https​://doi.org/10.1172/jci11​9771 (1997).
37. van Wout, A. B. et al. Efficient inhibition of both syncytium-inducing and non-syncytium-inducing wild-type HIV-1 by lamivudine
in vivo. AIDS 12, 1169–1176. https​://doi.org/10.1097/00002​030-19981​0000-00009​ (1998).
38. Cheng, C. et al. Genetic characteristics of HIV-1 CRF01-AE strains in four provinces, southern China. Chin. J. Epidemiol. 030,
720–725. https​://doi.org/10.3760/cma.j.issn.0254-6450.2009.07.019 (2009).
39. Cao, Z. et al. Trend of genetic subtypes and comparison of first CD4+T cell counts in newly diagnosed HIV infections in Liuzhou,
Guangxi, from 1998 to 2012. China J. Epidemiol. 40, 580–584. https​://doi.org/10.3760/cma.j.issn.0254-6450.2019.05.017 (2019).
40. Zhang, F. et al. Five-year outcomes of the China national free antiretroviral treatment program. Ann. Intern. Med. 151, 241–251.
https​://doi.org/10.7326/0003-4819-151-4-20090​8180-00006​ (2009).
41. Zhao, Y. et al. Mortality and treatment outcomes of China’s National Pediatric antiretroviral therapy program. Clin. Infect. Dis. 56,
735–744. https​://doi.org/10.1093/cid/cis94​1 (2013).
42. Zhang, F. et al. Effect of earlier initiation of antiretroviral treatment and increased treatment coverage on HIV-related mortality
in China: A national observational cohort study. Lancet Infect. Dis. 11, 516–524. https​://doi.org/10.1016/S1473​-3099(11)70097​-4
(2011).
43. National center for AIDS & STD Control and Prevention CC. Manual of the national free antiretroviral treatment (People’s Medical
Publishing House, Beijing, 2016).
44. Zhang, J. et al. Genetic characteristics of CRF01_AE among newly diagnosed HIV-1-infected 16- to 25-year olds in 3 geographic
regions of Guangxi, China. Medicine 94, e894. https​://doi.org/10.1097/MD.00000​00000​00089​4 (2015).
45. Zhong, P. et al. Genetic diversity and drug resistance of human immunodeficiency virus type 1 (HIV-1) strains circulating in
Shanghai. AIDS Res. Hum. Retroviruses 23, 847–856. https​://doi.org/10.1089/aid.2006.0196 (2007).

Scientific Reports | (2020) 10:15875 | https://doi.org/10.1038/s41598-020-72701-4 9

Vol.:(0123456789)
 www.nature.com/scientificreports/

46. Price, M. N., Dehal, P. S. & Arkin, A. P. FastTree: Computing large minimum evolution trees with profiles instead of a distance
matrix. Mol. Biol. Evol. 26, 1641–1650. https​://doi.org/10.1093/molbe​v/msp07​7 (2009).
47. Bender, R. Introduction to the use of regression models in epidemiology. Methods Mol. Biol. 471, 179–195. https​://doi.
org/10.1007/978-1-59745​-416-2_9 (2009).
48. Koenker, R. & Bassett, G. J. regression quantiles. Econometrica 46, 33–50 (1978).

Acknowledgements
Data in this manuscript were collected by Guangxi Center for Disease Control and Prevention.

Author contributions
Z.C., H.X., L.L. and Y.R. were responsible for study design and planning. J.L., H.C., C.S., Z.S., X.Z., G.L., Q.Z.
and S.L. contributed to data collection and management. Z.C., Y.F. and Y.R. contributed to data analysis. Z.C.,
H.X., L.L., Y.F., Y.S. and Y.R. contributed to interpretation. Z.C. and Y.R. contributed to writing the report. All
authors read and approved the final version of the paper.

Funding
This study was supported by the National Natural Science Foundation of China (Grants 11971479), the Ministry
of Science and Technology of China (2018ZX10721102-006, 2018ZX10715008), Guangxi Bagui Honor Scholars,
Guangxi Natural Science Foundation Project (Grants 2020GXNSFAA159020), and Chinese State Key Laboratory
for Infectious Disease Develop Grant.

Competing interests
The authors declare no competing interests.

Additional information
Correspondence and requests for materials should be addressed to Y.R.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the
Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this licence, visit http://creat​iveco​mmons​.org/licen​ses/by/4.0/.

© The Author(s) 2020

Scientific Reports | (2020) 10:15875 | https://doi.org/10.1038/s41598-020-72701-4 10

Vol:.(1234567890)