If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Institute of Molecular Biology and Biotechnology, National Institutes of Health, Manila 1000, PhilippinesInfectious Disease Section, Philippine General Hospital, Manila 1000, Philippines
Corresponding author at: Institute of Molecular Biology and Biotechnology, Room 108 National Institutes of Health, 623 Pedro Gil St, Manila 1000, Philippines. Tel.: +632182659.
The molecular epidemiology of HIV in the Philippines has changed substantially.
•
There is a unique opportunity for comparing outcomes and subtype characteristics between B and non-B subtypes.
•
There is evidence of faster progression of disease in CRF01_AE versus B, as represented by lower CD4 counts at presentation.
Abstract
Background
The Philippines has one of the fastest-growing HIV epidemics in the world. Possible reasons for this include increased testing, increased local transmission, and possibly more aggressive strains of HIV. This study sought to determine whether local molecular subtypes of HIV have changed.
Methods
Viruses from 81 newly diagnosed, treatment-naive HIV patients were genotyped using protease and reverse transcriptase genes. Demographic characteristics and CD4 count data were collected.
Results
The cohort had an average age of 29 years (range 19–51 years), CD4+ count of 255 cells/mm3 (range 2–744 cells/mm3), and self-reported acquisition time of 2.42 years (range 0.17–8.17 years). All were male, including 79 men who have sex with men (MSM). The genotype distribution was 77% CRF01_AE, 22% B, and 1% C. Previous data from 1985–2000 showed that most Philippine HIV infections were caused by subtype B (71%, n = 100), followed by subtype CRF01_AE (20%). Comparison with the present cohort showed a significant shift in subtype (p< 0.0001). Comparison between CRF01_AE and B showed a lower CD4+ count (230 vs. 350 cells/mm3, p = 0.03). Survival data showed highly significant survival associated with antiretroviral (ARV) treatment (p < 0.0001), but no significant difference in mortality or CD4 count increase on ARVs between subtypes.
Conclusions
The molecular epidemiology of HIV in the Philippines has changed, with the more aggressive CRF01_AE now being the predominant subtype.
The ongoing HIV/AIDS pandemic has been one of the world’s worst, with 60 million people infected since the beginning of the pandemic and 25 million deaths. HIV prevalence has remained fairly low in Southeast Asia, with Thailand and Cambodia the only countries with a prevalence above 1% (
). However, it is one of only nine countries in the world where the prevalence has increased more than 25% over the past decade, and annual reported cases have increased exponentially from 199 annual cases in 2004 to 7829 annual cases in 2015 (
The local epidemic has been characterized by an increase in local transmission vis-à-vis imported cases, and a shift from a heterosexual epidemic to a predominantly men who have sex with men (MSM) epidemic (
Molecular epidemiology of HIV-1 infection in the Philippines, 1985 to 1997: transmission of subtypes B and E and potential emergence of subtypes C and F.
). Two early studies in 1998 found that the majority of HIV infections were caused by subtypes B and CRF01_AE, with low numbers of subtypes A, C, D, and F (
). A shift from subtype B to CRF01_AE was shown after 1993 among female sex workers, and this was attributed to the closure of the United States military bases in the country as a predominant source of HIV among this group (
Molecular epidemiology of HIV-1 infection in the Philippines, 1985 to 1997: transmission of subtypes B and E and potential emergence of subtypes C and F.
). A recent study looking at infections between 2007 to 2012 (n = 163) showed that the predominant subtype remained B at 54%, followed by CRF01_AE at 37% (
). However, the majority of these samples were from people who inject drugs. A sub-analysis of the MSM population (n = 63) in this study revealed 64% CRF01_AE and 27% B.
With the large increase in HIV cases seen in the country, this study aimed to provide information regarding the genotypic characteristics of HIV in the Philippines, and to correlate this with available demographic and clinical data. Knowledge about the prevalence of different HIV subtypes among newly diagnosed cases of HIV infection will help determine transmission dynamics, inform treatment recommendations, and provide evidence for policy development.
Methods
Patient population
This study was approved by the University of the Philippines Research Ethics Board. Informed consent was obtained from all patients included in the study. This study was based at the Philippine General Hospital (PGH) HIV clinic in Manila. All active HIV-infected patients with confirmed Western blots not previously on antiretrovirals (ARVs) were recruited into this study. Written informed consent was obtained from the patients by clinic personnel for a blood draw of approximately 10 ml. Demographic, clinical, and laboratory data were collected from the patient charts by the clinic personnel, and the samples were de-identified using code numbers.
Inclusion and exclusion criteria
All adult HIV patients not currently on ARVs at the Philippine General Hospital were asked to participate. Exclusion criteria include refusal to participate and an indeterminate Western blot result.
HIV-1 drug resistance/subtyping PCR
RNA amplification: The protocol used for drug resistance testing was adapted from the World Health Organization (WHO) protocol for dried blood spots (
). The pol gene of HIV was reverse-transcribed in two overlapping segments using the Qiagen One-Step RT-PCR kit and custom primers. cDNA was synthesized from 3 μl of RNA, 5 × Reaction Mix, forward and reverse primers, Sensiscript RT/HotStar Taq Mix, and nuclease-free water in a 25-μl reaction mixture (Qiagen, Germany). For the reverse transcription, this mixture was incubated for 40 min at 50 °C. The amplification profile was as follows: initial cycle of 95 °C for 15 min; 35 cycles of 95 °C for 30 s, 53 °C for 30 s, 72 °C for 2.5 min; and a final extension of 10 min at 72 °C. Following the reverse transcription, the RT and PR regions of the pol gene were amplified using 5 μl from the first PCR, 10 × Reaction Mix, forward and reverse primers, Go Taq, and nuclease-free water in a 50-μl reaction mixture (Promega, Wisconsin, USA). The amplification profile was as follows: initial cycle of 95 °C for 10 min; 35 cycles of 94 °C for 20 s, 53 °C for 30 s, 72 °C for 2.5 min; and a final extension of 10 min at 72 °C. PCR products in the second run were analyzed using gel electrophoresis. The PCR products from the above protocol were shipped to a core facility (Macrogen, South Korea) for sequencing.
Sequence processing and analysis
Analysis of the subtypes and drug resistance mutations was done using the HIVDB program publicly available from the Stanford University website (
Automated subtyping of HIV-1 genetic sequences for clinical and surveillance purposes: Performance evaluation of the new REGA version 3 and seven other tools.
For phylogenetic and clustering analyses, sequences were aligned using MAFFT and trimmed to cover positions 1–1317 of HIV-1 PR/RT (HXB2 numbering). Reference sequences from the Los Alamos HIV Sequence Database (http://www.hiv.lanl.gov) spanning the same region, 1000 base pairs or longer, were downloaded and aligned to HXB2 PR/RT using bealign (https://github.com/veg/BioExt). Reference sequences were excluded if country or year of sampling were missing, resulting in a total of 104 127 references. Similar (<2% divergence) reference sequences were identified using TN93 (https://github.com/spond/TN93) as implemented in BioExt. Phylogenetic reconstruction of the combined dataset was performed using IQTREE v.1.5.3 (
) assuming a GTR + G4 + I model of substitution and rate variation.
Accession numbers
The sequences have been deposited in GenBank under accession numbers KX184117–KX184197.
Results
A total of 81 patients were recruited from March to August 2013. The cohort had an average age of 29 years (range 19–51 years), CD4 count of 255 cells/mm3 (range 2–744 cells/mm3), and self-reported acquisition time of 2.42 years (range 0.17–8.17 years). All were male, and 79 were MSM. Other baseline demographic characteristics of the sample are shown in Table 1.
Table 1Baseline demographic characteristics of the sample (N = 81).
Characteristics
Values
Comments
Age (years)
Mean 28.8, median 28 (SD 5.76), range 19–51
Sex
All male
CD4 count (cells/mm3)
Mean 255, median 228 (SD 205.21), range 2–744
Education
69 College level or higher, 12 high school, 1 elementary school
Occupation
34 (42%) unemployed, 10 (12%) call center workers, 37 (46%) other
Monthly income
Mean (with job n = 47) PHP 22 333 (SD PHP 12 097) Overall mean (N = 81) PHP 20 042 (SD PHP 12 729)
34/81 (42%) with previous STI; most common: urethritis/gonorrhea, anal warts, syphilis, some with multiple and concurrent infections
Previous HIV test
26/81 (32%) with previous negative HIV test
Anal sex
64 anal receptive, 60 anal insertive, 59 both
Condom use
8 never, 24 used <50% of the time, 36 used 50–75% of the time, 13 used >75% of the time
Healthcare worker
12/80 (15%); 5/12 (42%) admitted to accidental occupational needle-stick injury
1 with no response
Received money for sex
15/80 (19%)
1 with no response
Tattoo
15/80 (19%)
1 with no response
Sex with known IV drug user
17/80 (21%)
1 with no response
Sex with female sex worker
8/80 (10%)
1 with no response
Sex with person with known HIV/AIDS infection
19/80 (24%)
1 with no response
Sex with a foreigner
36/81 (44%)
CMV, cytomegalovirus; IV, intravenous; MDMA, 3,4-methylenedioxy-methamphetamine; MSM, men who have sex with men; OI, opportunistic infection; PCP, Pneumocystis pneumonia; SD, standard deviation; STI, sexually transmitted infection.
The genotype distribution showed 62 (77%) subtype CRF01_AE, followed by 18 (22%) subtype B and one (1%) subtype C. Three viral specimens exhibited some ambiguity in genotyping (all CRF01_AE), but were deemed to be CRF01_AE; this did not significantly affect any of the results when they were excluded from the analysis. One sample had possible transmitted drug resistance (M184V) conferring resistance against lamivudine in a patient with the CRF01_AE subtype. No other samples had drug resistance mutations.
Characteristics were compared between subtype CRF01_AE and subtype B, and it was found that CD4 counts were significantly lower in those with CRF01_AE than in those with B (230 cells/mm3 vs. 350 cells/mm3, p = 0.034 by t-test). This association persisted in a multivariate analysis, in which CD4 counts were square-root transformed and a linear model was fitted with subtype as a covariate (p = 0.0164), controlling for age and time to acquisition. In this model, there was no significant association of CD4+ count with age (p = 0.212) or time to acquisition (p = 0.798). Significantly higher rates of hepatitis B co-infection, illicit drug use, previous HIV testing, and higher number of lifetime partners were also found in those with the B subtype. Age, sex, and factors shown to be significant on univariate analysis are shown in Table 2. All other factors described in Table 1 were not significantly different between the subtypes (p> 0.05).
Table 2Comparison of selected parameters between subtypes CRF01_AE and B; t-test or Kruskal–Wallis test for continuous variables, Chi-square test for proportions.
Available outcome data for the entire cohort are summarized in Table 3. Three survival analyses were done: an intention to treat (ITT) analysis, where all transfers and lost to follow-ups were considered mortalities; a modified intention to treat (MITT) analysis, where all lost to follow-ups were considered mortalities but transfers were excluded; and a non-ITT analysis, where all those lost to follow-up and transfers were excluded. The average follow-up time was 1.5 years (range 0.44–2.36 years). Most patients were started on ARVs, and there was a highly significant increase in CD4 count and decrease in mortality across all analyses compared to those patients not started on ARVs.
Table 3Outcome data for the entire cohort.
Variable
Values
Comments
Outcome
60 alive, 6 dead, 5 lost to follow-up, 10 transfers
Mortality
26% (ITT), 15% (MITT), 10% (non-ITT)
Started on ARVs
65 (80.2%)
One patient with M184 V transmitted drug resistance started on rLPV
Subgroup analyses of subtype B vs. CRF01_AE looking at CD4 count increases or decreases with those on ARVs and mortality data are shown in Table 4. There was no significant difference in mortality between the two subtypes across all analyses, and there was no significant difference between CD4 increases in those on ARVs between subtypes. There was a statistically significant (p = 0.0305) higher rate of decline in CD4 count in patients with subtype B who were not started on ARVs compared to CRF01_AE, although these only represented three patients from each group with available data.
Table 4Outcome data comparing subtypes CRF01_AE and B.
Variable
CRF01_AE
B
Comments
Outcome
46 alive, 4 dead, 3 lost to follow-up, 9 transfers
13 alive, 2 dead, 2 lost to follow-up, 1 transfer
Average CD4 increase/year
Median 73.09 mm3/year
Median 22.35 mm3/year
p = 0.2555
On ARVs: median 76.96
On ARVs: median 105.82
p = 0.7535
Not on ARVs: median −13.07
Not on ARVs: median −117.00
p = 0.04953
ARV vs. no ARV outcome
ITT
ITT
ARV alive 44/50
ARV alive 12/14
No ARV alive 2/12
No ARV alive 1/4
p< 0.0001
p = 0.01681
MITT
MITT
ARV alive 44/45
ARV alive 12/13
No ARV alive 2/8
No ARV alive 1/4
p< 0.0001
p = 0.005517
Non-ITT
Non-ITT
ARV alive 44/45
ARV alive 12/13
No ARV alive 2/5
No ARV alive 1/2
p< 0.0001
p = 0.1013
ARV, antiretroviral; ITT, intention to treat analysis; MITT, modified intention to treat analysis.
As HIV cases continue to increase at an unprecedented rate in the Philippines, in stark contrast to the global trend, the use of molecular tools in conjunction with clinical and demographic data is essential in providing insights into the underlying dynamics of the epidemic. The present cohort, while small, is quite robust in that it was gathered prospectively at a national referral and training center for HIV and one of the largest treatment hubs for HIV in the Philippines.
The cohort provides several insights as analyzed. Pooled analysis of published data from samples collected from 1985 to 2000 showed that the majority of Philippine HIV infections were previously caused by subtype B (71%, n = 100), followed by subtype CRF01_AE (20%) and other non-B subtypes (9%). Comparison with the present cohort (B 22%, CRF01_AE 77%, C 1%; n = 81) showed a highly significant shift in subtype (p< 0.0001, by McNemar test). Comparison with a more recent cohort (B 54%, CRF01_AE 37%, other non-B subtypes 9%, n = 163) (
) from 2007–2012 likewise showed a statistically significant shift from B to CRF01_AE (p< 0.0001). The change in subtypes from 1985 to 2013 is shown in Figure 1, and a phylogenetic tree showing the relationship between the viral strains in this cohort and published sequences is shown in the Supplementary Material (Figure S1).
Figure 1Distribution of HIV-1 subtypes in the Philippines, 1985–2013.
Most of the B subtypes in previous studies were from people who inject drugs, which is not the predominant population in the present cohort, nor is it the predominant mode of transmission as reported by national data. Rather the present cohort is more similar to national statistics in that it is represented by a majority of MSM infected by sexual transmission. The subgroup of MSM samples (n = 63) from the 2007–2012 cohort (B 27%, CRF01_AE 63%, other non-B subtypes 10%) shows a more similar distribution to the present cohort, but is still statistically different (p = 0.0005), thereby confirming the continuing rise of CRF01_AE as the predominant strain in the MSM population.
This cohort itself is very similar to the August 2013 (time of last enrollment) national profile (
) of the epidemic, with the same median age (28 years), mostly male (95% vs. 100%), a majority of MSM transmission (84% vs. 98%), and similar proportions of homosexuals (51% vs. 49%) and bisexuals (36% vs. 46%), making it a reasonable representation of the current molecular epidemiology profile compared to the other available Philippine cohorts in the literature.
The mean CD4 count for the entire cohort was relatively low at 255 cells/mm3 (median 228 cells/mm3), with 34 patients presenting a CD4 count of less than 200 cells/mm3. Other characteristics are a high level of college education (85%), with a high unemployment rate (42%) despite a high level of educational attainment. This is not likely due to illness since 75% of the patients reported being in good health. There was a high rate of tuberculosis (38%), which is consistent with the high tuberculosis burden in the country and the active tuberculosis rates in the authors’ HIV clinic, with respect to the low median CD4 counts. Hepatitis B co-infection is consistent with that previously reported from the authors’ clinic, at 11% (
). There were high rates of opportunistic co-infections as well, consistent with the low CD4 counts at presentation. The apparent disparity between the short self-reported time to acquisition and the low CD4 count may be due to limited recall, accelerated time to AIDS for both subtypes due to concurrent tuberculosis, or an intrinsic property of HIV subtypes in the Philippines.
There was a high rate of tobacco use in this patient cohort, with over 50% having smoked, representing a further health risk. Illicit drug use was present in about a third of patients, with four reporting intravenous drug use, including one with intravenous methamphetamines. It is difficult to definitively assign the mode of acquisition of the four patients who also used intravenous drugs, but going by the genotypes (one B, three CRF01_AE), at least one of them may have become infected via contaminated needles first.
The average age at first sex was before 18 years, with a median of 20 partners, which is quite concerning as the epidemic continues to grow, since there are many opportunities for transmission. Most (n = 79) were MSM, but nearly half (47%) also had female partners who they may transmit to. Due to the concentrated nature of the epidemic, female partners of these men may be unaware of their high risk, and this may further delay testing and health-seeking behavior. Nearly half (42%) reported previous sexually transmitted infections (STIs) and this is concerning due to the increased risk for transmission and progression between STIs and HIV (
). Only a third of this cohort had previously accessed an HIV test, belying the poor awareness and/or reluctance to test in the midst of the growing HIV problem. While most (90%) had used condoms, only 16% used these more than 75% of the time.
A significant proportion of patients were healthcare workers (15%), with 42% of them admitting to occupational needle-stick exposure. While it is unlikely that HIV was acquired through this mode given prevailing risk factors, hospital infection control units need to be aware that needle-stick exposure is not uncommon, and the infection risk may increase as the prevalence of HIV in the population increases. Nearly a quarter of this cohort had had sex with a person with a known HIV infection, in addition to nearly half (44%) having had sex with a foreigner, highlighting the multiple opportunities for the epidemic to grow in the Philippines.
A comparison between subtype B and subtype CRF01_AE showed lower CD4 counts at presentation for the latter, which is concerning due to the emergence of CRF01_AE as the predominant genotype (Figure 2).
), and this finding is consistent not only with the present cohort, but also the increasing progress of the epidemic, as the predominant subtype has shifted.
The finding of a higher hepatitis B co-infection rate, illicit drug use, and larger number of sexual partners in the subtype B group is plausible, since subtype B is more predominant in drug users, and a larger number of partners would predispose to higher hepatitis B co-infection rates.
Out of the original cohort, 60 (74%) were alive at the last follow-up time. A mortality of 0–26% was found using the different analyses, depending on how those lost to follow-up were handled. Most patients were started on ARVs, specifically the generic 3-in-1 efavirenz/tenofovir/lamivudine WHO regimen, and showed a substantial degree of immune recovery, as well as a significantly lower mortality compared to those who did not start ARVs. This increased survival was significant across all analyses, and is further evidence of the effectiveness of ARVs in Filipinos.
There were no significant differences in outcomes between subtypes. This is not surprising given the small sample sizes and the short period of observation. Patients with either subtype showed substantial immune reconstitution on ARVs across all analyses. There was a finding of a faster decline in CD4 count for subtype B in those not started on ARVs, but this is difficult to interpret since the number of patients involved in this analysis was only six, three for each genotype.
It was demonstrated in this study that the molecular epidemiology of HIV in the Philippines has shifted substantially. The change in subtypes is particularly significant given the unprecedented increase in HIV cases in the Philippines, and the change in demographics from heterosexual to MSM transmission. Due to the paucity of data on the natural history of non-B subtypes, the epidemiological shift in subtypes from a B to a non-B subtype within a country represents a unique opportunity for comparing outcomes and subtype characteristics. In this small cohort alone, some evidence of faster progression of disease in CRF01_AE versus B was seen, as represented by lower CD4 counts at presentation. In addition, despite the small sample size and the presumably long time since infection, five pairs of individuals had similar (<1% divergence) sequences; such clustering is suggestive of a high transmission rate. The high number of sexual partners coupled with poor condom use will ensure that this epidemic continues to grow unless drastic measures are taken. The substantial immune reconstitution following initiation of ARVs in both subtypes is somewhat reassuring, but this needs to be observed for longer periods of time, especially with regard to the propensity for the development of drug resistance in a circulating recombinant form, and the possibility of immunological failure. The recently added WHO recommendation of pre-exposure prophylaxis is especially relevant in the context of possible emergence of resistance, as these interventions have mostly been studied in subtype B. It is recommended that studies with larger cohorts and longer follow-up times are conducted to explore these phenomena.
Funding
This work was supported by a grant from the Philippine Council for Health Research and Development of the Department of Science and Technology. SDWF was supported in part by an MRC Methodology Research Programme grant (MR/J013862/1).
Conflict of interest
The authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers’ bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript.
Appendix A. Supplementary data
The following is Supplementary data to this article:
Automated subtyping of HIV-1 genetic sequences for clinical and surveillance purposes: Performance evaluation of the new REGA version 3 and seven other tools.
Molecular epidemiology of HIV-1 infection in the Philippines, 1985 to 1997: transmission of subtypes B and E and potential emergence of subtypes C and F.
30154-6&id=gr1.jpg){kind=link}
30154-6&id=gr2.jpg){kind=link}