Comparison of Plasma,Dried Blood Spots,and Peripheral Blood Mononuclear Cells as Biosamples for HIV-1 Genotypic Drug Resistance in a Tertiary Care Center

Abstract

The collection, storage, and transport of plasma, the ideal specimen for HIV-1 genotyping, is plagued by technical difficulties in resource-limited settings. We aimed to compare corresponding bio-samples for HIV-1 genotypic drug resistance testing. A total of 87 matched specimens of plasma, dried blood spots (DBS), and peripheral blood mononuclear cells (PBMCs) collected from 29 persons living with HIV (PLWH) in clinical, immunological, and/or virological failure were included. Drug resistance genotyping was done by nested PCR amplification and Sanger sequencing of the HIV-1 pol gene. The clinical reporting was based on the Stanford University HIV Drug Resistance Database. Amplification and genotyping success rates from the three sample types were compared. The level of agreement between the sample types was assessed using Cohen’s kappa coefficient. In total, 89.7% (n = 26) of samples were amplified in plasma, 69% (n = 20) in DBS, and 100% (n = 29) in PBMC. In samples with plasma viral load >1,000 copies/mL, 96.2% were amplified in plasma, 73.1% in DBS, and 100% in PBMCs. The median number of mutations detected in plasma, DBS, and PBMCs was 6.5 (interquartile range [IQR]: 2–8.25), 5 (IQR: 0–6), and 5 (IQR: 2–7), respectively. The difference in the number of mutations across the three sample types was not statistically significant (p = 0.221). The agreement between the sample types was calculated based on susceptibility and resistance to different antivirals. The kappa values for nucleoside reverse transcriptase inhibitors and non-nucleoside reverse transcriptase inhibitors ranged from 0.70 to 0.88 and 0.75 to 0.87, respectively. Six samples showed discordance in HIV-1 drug resistance profiles when compared across the three compartments. DBS is a promising alternative to plasma for HIV-1 genotypic testing in resource-limited settings owing to the ease of sampling, storage, transportation, human resource efficiency, and cost-effectiveness. However, no single specimen type can satisfy all requirements and purposes. Selecting an appropriate specimen for a setting requires careful consideration of the practical constraints, logistical capacity, and application needs.

Keywords

dried blood spots DBS ambient temperature HIV drug resistance genotyping India

Introduction

Antiretroviral therapy (ART) has transformed the management and prognosis for people living with human immunodeficiency virus (HIV) infection. Nevertheless, an HIV infection still demands lifelong treatment as it continues to be incurable. As a result of many years of treatment, drug resistance mutations (DRMs) against the antiretroviral agents can develop due to poor adherence and suboptimal regimens.¹

Drug resistance genotyping is critical in the management of people living with HIV (PLWH), as it directs treatment decisions and the timing and choice of regimen switches.² Selecting an appropriate therapeutic regimen, achieving virological suppression, and curtailing transmission benefits both the individual and the community. In addition, the data generated from genotyping will provide insight into the extent and prevalence of drug resistance among PLWH, particularly in resource-limited settings, which can, in turn, help shape clinical guidelines for second- and third-line treatment regimens.

Genotypic assays are the preferred method for HIV-1 drug resistance testing.³ Plasma is the most appropriate clinical sample⁴ since it is known to have HIV-1 RNA at higher and more stable levels than serum and whole blood.⁵ The samples for HIV-1 drug resistance testing should be collected while the individual is on the failing ART regimen. This ensures that the resistance mutations developed during treatment are sustained and detectable upon testing.⁶ Plasma must be separated from blood cells within 6 hours of collection to prevent RNA degradation and stored at −70°C until the time of testing.⁴ These genotypic assays are available only in select laboratories in India, as they are not covered under the national program and are reserved for PLWH who fail second or third-line regimens. The lack of adequate facilities and equipment, the effort required to maintain the cold chain of plasma during transportation, and inadequate storage facilities leave resource-limited settings incapable of managing PLWH satisfactorily.⁷

The dried blood spot (DBS) is being studied as an alternative specimen to plasma. The ease of collection of whole blood samples onto a DBS card, its storage at room temperature, and transport to reference laboratories at ambient temperatures have made DBS an appealing sample for HIV-1 drug resistance testing.⁷

Many studies have reported the successful genotyping of HIV-1 from DBS, and some have shown a high genotypic concordance with plasma genotypes.^8–11 DBS has started to be used widely for HIV-1 drug resistance testing worldwide, and an increased number of reports from resource-limited areas have indicated DBS as the preferred specimen for the surveillance of HIV-1 transmitted drug resistance, where the collection of plasma is not feasible.¹² From India, though there is experience with HIV-1 viral load monitoring from DBS,¹³ there is only limited information available on HIV-1 drug resistance genotyping using DBS samples stored at ambient temperature.¹⁴ Another bio-sample explored for HIV-1 genotyping is peripheral blood mononuclear cells (PBMCs), which are known to harbour proviral DNA and function as reservoirs for archived mutations.^15,16

These reasons led to this study testing for DRMs in corresponding plasma, DBS cards, and PBMC samples, and evaluating the efficacy of DBS in detecting HIV-1 drug resistance and the concordance in the mutations in the three sample types. The study aimed to ascertain the genotyping success rates of DBS stored at ambient temperature, thereby assessing the feasibility of DBS as a cheaper and pragmatic alternative to plasma for HIV drug resistance genotyping in resource-constrained settings.

Materials and Methods

Study design, setting, and participants

This cross-sectional study was approved by the Institutional Review Board (IRB Min. No. 9832, dated 07.01.2016) and conducted between 2016 and 2021 at a tertiary care center in southern India. Participants were recruited from July 2016 to July 2017, after obtaining informed consent. The inclusion criteria were as follows: 18 years of age or older, have a serologically confirmed HIV infection, and be ART-experienced PLWH in treatment failure. Consecutive PLWH referred to the Department of Virology for HIV-1 drug resistance genotyping, due to clinical, immunological, and/or virological failure, were approached to enroll in the study. Matched specimens of plasma, DBS, and PBMCs were collected from consenting PLWH.

Sample collection and preparation

After obtaining informed consent, 8 mL of blood was collected in a sterile Ethylenediaminetetraacetic acid-containing tube for routine HIV-1 drug resistance testing; no additional blood was collected for the study. Five spots, each of 50 µL of whole blood, were spotted on a Whatman 903 filter paper card,¹⁷ which was left to dry overnight in the biosafety cabinet. The next day, the blood spots were covered with glassine paper and packed in a plastic zip-lock pouch along with two desiccants. The pouch was kept at 25–30°C for 10 days and then stored at −20°C until testing. For extraction, three 6 mm blood spot punches were taken using a single-hole, hand-held punch. The punch was wiped with 95% ethanol between each sample to prevent carry-over. After spotting, the remaining blood was centrifuged at 2,000 rpm for 10 minutes at 4°C. The plasma was separated, and multiple aliquots were made and stored at −70°C until testing as part of the standard of care.¹⁸ After plasma was separated and aliquoted, PBMCs were extracted from the remaining blood using Ficoll-Paque® (GE Healthcare, Uppsala, Sweden) density gradient centrifugation and stored at −70°C until testing.

Estimation of HIV-1 viral load

HIV-1 viral load testing in plasma samples was performed using the Abbott RealTime HIV-1 assay (Abbott, Abbott Park, Illinois, USA) on the Abbott m2000 RealTime System, using the 0.6 mL protocol with lower and upper limits of quantification of 40 copies/mL (log 1.60) and 10,000,000 copies/mL (log 7.0), respectively. HIV-1 viral load testing in DBS samples was done as part of troubleshooting for those samples with plasma amplification, but with corresponding DBS cards that did not amplify. The DBS HIV-1 viral load testing was performed using the Abbott RealTime HIV-1 assay (Abbott, Abbott Park, Illinois, USA), following a previously standardized modified protocol from this center.¹⁹ The results of Abbott HIV-1 viral load testing were interpreted as “Target not detected” (TND), <40 copies/mL, or actual copy numbers/mL.

PCR amplification of PR and RT genes and HIV drug resistance genotyping

Extraction of nucleic acids from plasma, DBS, and PBMCs was performed manually, using QIAamp® viral RNA extraction kit, QIAamp® DNA Mini extraction kit, and QIAamp® DNA Blood Mini extraction kit (Qiagen GmbH, Hilden, Germany), respectively.

HIV-1 drug resistance genotyping from plasma was performed using conventional nested PCR amplification and Sanger sequencing of the reverse transcriptase (RT) and protease (PR) regions of the pol gene. Three forward and three reverse overlapping primers were utilized for bidirectional sequencing of the amplified product in six separate reactions, employing the BigDye Terminator assay with the Applied Biosystems 3500 genetic analyzer (CA, USA). The sequences were visualized using the FinchTV chromatogram viewer, aligned, and a consensus of the forward and reverse reads was assembled using BioEdit Sequence Alignment Editor. The consensus was then submitted to the Stanford HIV Drug Resistance Database (HIVDB) for interpretation (https://hivdb.stanford.edu/).^18,20

For resistance genotyping from DBS and PBMCs, the HIV-1 pol gene was amplified using a nested PCR to produce a final product of 1,800 base pairs. The cycling condition for the first round of PCR was as follows: 95°C for 15 minutes for activation of Hot Start Taq polymerase enzyme, followed by 45 cycles of 94°C for 45 seconds, 55°C for 45 seconds, and 72°C for 2 minutes, followed by a final extension at 72°C for 7 minutes and then held at 4°C. The second round of PCR had the following cycling conditions: 95°C for 15 minutes for activation of the enzyme, followed by 40 cycles of 94°C for 45 seconds, 55°C for 45 seconds, and 72°C for 2 minutes, followed by a final extension at 72°C for 7 minutes and then held at 4°C. Gel electrophoresis and documentation were used to assess the amplification success by checking for the specific 1,800 bp amplicon of the PR and RT genes.

The subsequent sequencing of the amplified products, alignment, analysis, and interpretation of sequences were performed following the same protocol as described above for plasma specimens.¹⁸

The clinical reporting of HIV-1 drug resistance genotyping was based on the Stanford HIVDB.²¹ The HIVDB Genotypic Resistance Test (GRT) Interpretation System uses DRM penalty scores. The overall activity of antiretrovirals is estimated by adding the penalties for each DRM and classifying drug resistance into five different levels: (i) Susceptible, <10; (ii) Potential low-level resistance, 10–14; (iii) Low-level resistance, 15–29; (iv) Intermediate resistance, 30–59; and (v) High-level resistance, ≥60. For reporting the study samples, we used a DRM penalty score cutoff of 15 to categorize each class of antiretroviral, that is, nucleoside reverse transcriptase inhibitors (NRTIs), non-nucleoside reverse transcriptase inhibitors (NNRTIs), and protease inhibitors (PIs), into two levels, as “Susceptible” (<15) or “Resistant” (≥15), for each sample type. If the DRMs identified in any two sample types were the same or had minor differences that did not lead to a difference in drug resistance level, then those samples were considered to have identical drug resistance profiles.

In this study, if the genotyping results (drug resistance profile) from any two sample types of an individual were in agreement, that is, if both were susceptible or were resistant, those samples were considered “concordant.” If the drug resistance profile of a sample varied across any pair of compartments, then those two sample types were taken as “discordant.”

Modifications to the protocol for DBS samples, which did not amplify

For troubleshooting DBS samples that did not amplify, the following modifications were attempted during the extraction and amplification of nucleic acids: the input for extraction was increased, the volume of extract used was increased for first round PCR, and the volume of the first round product was increased for the second round of PCR, different primer sets, and RNA-specific extraction and amplification was also tried. Finally, we also estimated the HIV-1 viral load directly from the DBS cards, which did not amplify.

Statistical analysis

The mean with standard deviation or the median with interquartile range (IQR) was calculated for the continuous variables. Frequencies with percentages were calculated for categorical variables. Sensitivity was calculated for each test using plasma as the gold standard. Observed percentage agreement was calculated for each pair of sample types. The level of agreement between the sample types was assessed using Cohen’s kappa coefficient. p value < .05 was considered statistically significant. GraphPad Prism Version 10.3.0 (GraphPad Software, Boston, MA, USA) was used for analysis.

Results

Baseline profile of the participants and their ART

A total of 87 matched specimens of plasma, DBS, and PBMCs were obtained from the 29 PLWH included in the study. Of the 29 study participants, 79% (n = 23) were male. The age of the participants ranged from 18 to 56 years, with a median of 46 years (IQR = 39–49 years). The participants hailed from the following Indian states: Tamil Nadu (n = 13, 44.8%), Andhra Pradesh (n = 12, 41.4%), Kerala (n = 2, 6.9%), Bihar (n = 1, 3.4%), and Jharkhand (n = 1, 3.4%). The virological and treatment profiles of the participants are described in Table 1.

Table 1.

The Virological and Treatment Profiles of the Participants

Parameters	Findings
Profile of study participants
CD4⁺ T cell count
Range	5–775 cells/μL
Median age (IQR)	155 (68–319) cells/μL
Plasma viral load (VL) levels
Range	615–1,735,708 copies/mL
Median VL (IQR)	58,242 (8,842–195,236) copies/mL
Median VL (IQR) in log	log 4.77 (3.74–5.34) copies/mL
>1,000 copies/mL (>3 log copies/mL)	n = 26 (89.7%)
<1,000 copies/mL (<3 log copies/mL)	n = 3 (10.3%)
Profile of antiretroviral therapy^a
Duration of ART
Range	0.25–16 years
Median duration (IQR)	4.5 (1.25–9) years

ART ^a regimen switch vs. ART duration
No. of ART switches	0	1	2	4
ART duration (years)
<1	2
1–5	9	4	2
6–10	4	2	1
11–15		1	1	1
>15			1
Total (%)	15 (53.6%)	7 (25%)	5 (17.9%)	1 (3.5%)

ART details were not available for one participant.

ART, antiretroviral therapy; IQR, interquartile range.

The 28 study participants followed 13 different ART regimens at the time of sample collection. The most frequently used regimens were Tenofovir + Emtricitabine + Efavirenz (n = 6, 21.43%) and Tenofovir + Lamivudine + Efavirenz (n = 5, 17.85%). Six individuals had discontinued their treatment and then restarted it, and three were poorly adherent to the therapy.

Amplification success rates of plasma, DBS, and PBMC specimens

In total, 89.7% (n = 26/29) of study samples were amplified in plasma, 69% (n = 20/29) in DBS, and 100% (n = 29/29) in PBMCs. The distribution of amplified samples among the three different compartments is depicted in Figure 1.

FIG. 1.

Venn diagram illustrating the number of samples that were amplified in plasma, DBS, and PBMC (n = 29). DBS, dried blood spots, PBMC, peripheral blood mononuclear cells.

From the 26 samples with plasma viral load >1,000 copies/mL, 25 (96.2%) amplified from the gold standard for HIV-1 genotyping, the plasma sample. From DBS and PBMCs, 19 (73.1%) and 26 (100%) samples were amplified, respectively. The plasma HIV-1 viral load of the amplified samples ranged from 3.02 to 6.24 log₁₀ copies/mL. The amplification success for HIV-1 genotyping in samples with a plasma HIV-1 viral load >1,000 copies/mL (n = 26) was 25 (96.2%) for plasma, 19 (73.1%) for DBS, and 26 (100%) for PBMCs, respectively, p = .005.

Among the remaining three samples with plasma HIV-1 viral load <1,000 copies/mL, 1 (33.3%) was amplified from plasma, 1 (33.3%) from DBS, and all 3 (100%) from PBMCs.

Five samples with HIV-1 plasma viral load of more than 4.5 log₁₀ copies/mL (Median = 5.27 log₁₀ copies/mL) did not amplify in DBS despite multiple repeat tests and modifications. The HIV-1 DBS viral load for two of these samples was TND, and three of the remaining samples had an HIV-1 DBS viral load of less than 3.8 log₁₀ copies/mL (Median = 2.63 log₁₀ copies/mL). On excluding these five samples, there were 21 samples with a plasma HIV-1 viral load >1000 copies/mL, and amplification success for plasma, DBS, and PBMCs were 95.23% (n = 20/21), 90.50% (n = 19/21), and 100% (n = 21/21), respectively.

Successful amplification of PR and RT genes in all three sample types—plasma, DBS, and PBMCs was demonstrated in 62.1% (n = 18/29) of samples; 27.6% (n = 8/29) samples amplified from plasma and PBMCs only, 6.9% (n = 2/29) samples amplified from DBS and PBMCs only; and one sample with 615 copies/mL had amplified only in PBMCs.

The overall sensitivity of DBS and PBMCs with respect to plasma for 29 samples was 69.2% and 100%, respectively.

Genotyping success rates of plasma, DBS, and PBMC specimens

Upon analysis, all sequenced genomes were found to belong to HIV-1 subtype C.

Plasma versus DBS

18 samples were amplified from both plasma and DBS, and sequences were available for 17 of these samples. One sample (DBS-024) was amplified from DBS; however, the sequence was unavailable for analysis.

Concordant samples

Amplified DBS and plasma samples were compared, and 71% (n = 12/17) were concordant. Analysis of their mutation profile revealed that four had no drug-resistance mutations, two had identical drug-resistance mutations, and two samples had different amino acids at the same mutation site. Four samples had additional mutations that affected the resistance level in certain drugs but did not affect the overall resistance profile. Among the two samples with identical mutation profiles in plasma and DBS, none had mutations that conferred resistance to PIs; all had conferred resistance to NRTIs and NNRTIs.

Discordant samples

Of the amplified paired plasma and DBS samples analyzed, 29% (n = 5/17) showed discordance between plasma and DBS. Analysis of the drug resistance profiles of these samples showed that the drug resistance profile varied in all three classes of antiretrovirals: PIs, NRTIs, and NNRTIs. Three samples (DBS-012, -016, -026), only had discordant drug resistance profiles in NNRTIs. One (DBS-003) was discordant in both NRTIs and NNRTIs, and the last (DBS-009) was discordant in both PIs and NRTIs. The mutations in the specimens with discordant drug resistance profiles are given in Table 2.

Table 2.

The DRMs Detected in Discordant Drug Resistance Specimens

Sample ID	Mutations in plasma			Mutations in DBS			Mutations in PBMCs
	PI	RT		PI	RT		PI	RT
		NRTI	NNRTI		NRTI	NNRTI		NRTI	NNRTI
DBS-003		None	M41L, D67N, K70R, L74I, M184V, T215F, K219Q		V106M, G190A, F227L, M230L	None		None	None	None	None	None
DBS-009	V32I, M46I, I54M, I84V, L90M	M41L, D67N, K70R, M184V, T215F, K219E	A98G, Y181C	I54V, V82A	M184V	A98G, Y181C	None	M41L, M184V, T215F	A98G, Y181C
DBS-012	None	None	None	None	None	V106M	None	None	None
DBS-016	None	None	K103N, G190A	None	None	K103N	None	None	K103N, G190A
DBS-022^a	V82A	D67N, T69D, K70R, M184V, K219E	L100I, K103N				None	D67N, T69D, K70R, M184V, K219E	L100I, K103N, N348I
DBS-026	None	K70E, M184V	K103N, V108I, H221Y	None	K70E, M184V	K103N, V108I, E138G, H221Y	None	K70E, M184V	K103N, V108I, H221Y
DBS-030	None	None	None	None	None	None	None	M184V	V106M

The differences in DRMs are indicated in bold.

The dried blood spot of DBS-022 did not amplify.

DRM, drug resistance mutations; PI, protease inhibitors; NRTI, nucleoside reverse transcriptase inhibitors; NNRTI, non-nucleoside reverse transcriptase inhibitors; RT, reverse transcriptase.

Plasma versus PBMCs

Among the samples amplified in plasma, 96.1% (n = 25/26) had an HIV-1 viral load above 1,000 copies/mL, and all amplified plasma samples had paired amplified PBMC samples.

Concordant samples

Among the amplified paired plasma (HIV-1 viral load >1,000 copies/mL) and PBMC samples, 84% (n = 21/25) showed concordance, having identical drug resistance profiles in both plasma and PBMCs. Analysis of their mutation profile revealed that 10 samples had identical DRMs, four had no DRMs, and seven had additional DRMs. Among the 10 samples with identical mutation profiles in plasma and PBMCs, one had mutations that conferred resistance to PIs, and all had conferred resistance to NRTIs and NNRTIs.

Discordant samples

Out of the amplified paired plasma and PBMC samples analyzed, 16% (n = 4/25) showed discordance between plasma and PBMC. Two samples (DBS-009, -022) showed discordant drug resistance profiles only in PIs, whereas the remaining two (DBS-003, -030) showed discordance in both NRTIs and NNRTIs (See Table 2).

DBS versus PBMC

Twenty samples were amplified from both DBS and PBMCs, and sequences were available for 19 of these samples.

Concordant samples

Amplified DBS and PBMCs were compared, and 74% (n = 14/19) were concordant. Analysis of their mutation profile revealed that four pairs had no drug-resistance mutations, three had identical drug-resistance mutations, two samples had different amino acids at the same mutation site, and five samples had additional drug-resistance mutations in the DBS or PBMCs, which did not affect the overall resistance profile.

Discordant samples

Out of the amplified paired DBS and PBMC samples analyzed, 26% (n = 5/19) showed discordance between DBS and PBMCs. Analysis of the drug resistance profiles of these samples showed that the drug resistance profile varied in all three classes of drugs. Three samples (DBS-012, -016, -026), only had discordant drug resistance profiles in NNRTIs. One (DBS-030) was discordant in both NRTIs and NNRTIs, and the last (DBS-009) was discordant in both PIs and NRTIs (See Table 2).

Plasma versus DBS versus PBMCs

Sequence data were available for all 3 compartments in 17 out of the 29 samples.

Concordant samples

The drug-resistant profile was concordant in all three compartments for 65% (n = 11/17) of the samples. Eight samples showed similar drug-resistance profiles in all three compartments, while three samples were susceptible in all three compartments.

Discordant samples

Six samples showed discordance when compared across three compartments. A summary of the patterns of discordance is presented in Table 3.

Table 3.

Discordance Pattern Between Plasma, DBS, and PBMCs

Study ID	Plasma versus DBS	Plasma versus PBMCs	DBS versus PBMCs	Plasma versus DBS versus PBMCs
DBS-003	Discordant	Discordant	Concordant	Discordant
DBS-009	Discordant	Discordant	Discordant	Discordant
DBS-012	Discordant	Concordant	Discordant	Discordant
DBS-016	Discordant	Concordant	Discordant	Discordant
DBS-026	Discordant	Concordant	Discordant	Discordant
DBS-030	Concordant	Discordant	Discordant	Discordant

The concordant results of pairs of biosamples are indicated in bold.

ART in discordant samples

Among the discordant samples, 67% (n = 4/6) of individuals experienced ART switches during their ART, with three individuals having a single ART change and one individual (DBS-009) having four ART switches (See Table 4). This individual (DBS-009) had been on ART for 14 years and had been put on different classes of antiretrovirals, which included NRTI, NNRTI, PI, and integrase strand transfer inhibitors (INSTIs). In DBS-009, mutations against NRTIs and NNRTIs were detected in all three sample types, and mutations against PIs were detected in plasma and DBS (See Table 2). The drug resistance profile of DBS-009 showed discordance between all three compartments (See Table 3).

Table 4.

Antiretroviral Therapy Duration, Number of Switches, and ART Regimens in Individuals with Discordant Results

Sample ID	ART duration	Switches	Regimen 1	Regimen 2	Regimen 3	Regimen 4	Regimen 5
DBS-003	4 years	1	ZDV+3TC+NVP (1 year)	ZDV+3TC+EFV (3 years)
DBS-009	14 years	4	d4T+3TC+EFV (4 years)	d4T+3TC+NVP (3 years)	TDF+FTC+LPV/r (2 years)	TDF+FTC+DRV/r (1 year)	ZDV+3TC+RAL+DRV/r (4 years)
DBS-012	10 years	1	TDF+FTC+EFV (5 years)	ABC+3TC+EFV (5 years)
DBS-016	2 years	1	ZDV+3TC+NVP (0.4 year)	TDF+3TC+NVP (2 years)
DBS-026	1 year	0	TDF+FTC+EFV (1 year)
DBS-030	3 years	0	ABC+3TC+EFV (3 years with irregular ART for first 2 years)

ABC, abacavir; d4T, stavudine; DRV/r, darunavir/ritonavir; EFV, efavirenz; FTC, emtricitabine; LPV/r, lopinavir/ritonavir; NVP, nevirapine; RAL, raltegravir; 3TC, lamivudine; TDF, tenofovir; ZDV, zidovudine.

The agreement of the HIV-1 genotypic assay between the sample types was calculated based on susceptibility and resistance to different antivirals. The kappa values for NRTIs and NNRTIs were 0.70–0.88 and 0.75–0.87, respectively. The kappa value for PIs was not calculated since only two samples had mutations to PIs.

The median number of mutations detected in plasma, DBS, and PBMCs was 6.5 (IQR: 2–8.25), 5 (IQR: 0–6), and 5 (IQR: 2–7), respectively. There was no statistically significant difference (p = 0.221) in the number of mutations detected across the three sample types. Upon stratifying the samples by a viral load of 5,000 copies/mL and comparing the proportion of concordant and discordant samples in each group, the difference was not statistically significant (p = 0.5). For the 20 samples that were amplified in DBS, the total number of major mutations detected in plasma, DBS, and PBMCs is depicted in Figure 2.

FIG. 2.

Scatter plot showing the total number of major mutations identified in plasma, DBS, and PBMC in the 20 samples that were amplified in DBS.

Discussion

Combination ART is not a cure for HIV infection but, it can effectively suppress viral replication and alter the natural course of the disease from a progressive condition with a fatal outcome into a chronic, manageable disease.²² However, there is a reasonable probability of developing drug resistance due to the high replication and mutation rate of the virus, the chronic nature of the infection, the necessity for lifelong treatment, and the lack of adherence. Therein arises the need for HIV drug resistance testing, both for the optimal management of PLWH and for surveillance purposes. To address the operational, practical, and economic difficulties associated with conventional plasma, the WHO recommends DBS as a field-friendly specimen that is highly suitable for low- and middle-income countries.^17,23

Most of the available data on HIV-1 drug resistance to ART are for subtype B HIV-1, which is prevalent in developed countries but is responsible for only about 12% of global infections. In comparison, non-B subtypes account for approximately 88% of all infections and are primarily seen in low- and middle-income countries.²⁴ As expected, in our study, all the infections were caused by HIV-1 subtype C, which is the predominant subtype seen in India.²⁵

Our cross-sectional study analyzed the genotyping success rates of matched plasma, DBS stored at ambient temperature, and PBMC specimens collected from treatment-experienced individuals using in-house amplification assays in a tertiary care hospital in a resource-limited setting in southern India. The overall amplification success rates for plasma, DBS, and PBMC specimens, irrespective of the viral load, were 89.66%, 68.97%, and 100%, respectively.

The rate of successful amplification and sequencing of HIV-1 RNA from plasma is directly proportional to the viral load in a sample.²⁶ In a study by Diallo et al., an amplification success rate of 87% was observed in plasma using a broadly sensitive in-house assay.²⁷ Zhou et al. studied non-B subtypes detected from treatment-experienced individuals’ plasma with viral load >3log₁₀ and <3 log₁₀ copies/mL and obtained 100% and 78.6% success rates, respectively.²⁸ Among our plasma samples, 96.2% (25/26) amplified when the viral load was ≥3 log₁₀ copies/mL, which is comparable to other studies. And when the Viral load (VL) was <3 log₁₀ copies/mL, there was poor amplification success, with only one-third of the samples amplifying.

Several studies have been conducted to evaluate the use of DBS for HIV-1 genotyping and have had varying degrees of success, ranging from 30%–100%.²⁹ RT alone or PR and RT gene amplification have been done using diverse testing methods. The storage conditions of the DBS cards, the characteristics of the study participants, the sample size, and the viral load of the tested samples have varied significantly across studies.²⁹ The reported amplification sensitivity from DBS markedly differs between assays, with some in-house tests having successful amplification from DBS at low viral loads.²⁹ All reports indicate a slightly lower sensitivity of DBS than plasma, due to the smaller input volume and viral RNA degradation.²⁶ In our study, the DBS cards stored at 25°−30°C for 10 days to simulate transport temperatures, and then at −20°C until testing, showed 73% amplification success in samples with VL >1,000 copies/mL, while Parry et al., had around 90% and 63% success rates in DBS cards kept at ambient temperature for 2 weeks and 4 weeks, respectively, when the viral load ranged from 1,000 to 10,000 copies/mL.³⁰ Neufeld et al. reported sensitivities of 68.9% (58.3%–78.2%) and 77.8% (67.8%–85.9%) for their in-house assay in amplifying PR and RT genes, respectively, at 1,000 copies/mL viral load. They tested DBS cards that were dried for 24 h at room temperature and then immediately stored at −80°C.³¹

Our study included five samples with high viral loads ranging from log₁₀ 4.98 to 5.79 copies/mL, which did not amplify in DBS despite troubleshooting and repeat testing. These cards were examined to confirm the integrity of the filter paper card and were found to be satisfactory. The presence of mutations at the primer binding site was considered. However, this was unlikely as the corresponding plasma and PBMC samples were amplified. Fewer punches were taken, and the extract was diluted and retested to rule out a higher nucleic acid concentration. A new batch of reagents was used for all repeat testing. HIV-1 viral load estimation was also performed from these DBS cards. All the packed DBS cards were stored in the same container at −20°C until testing; only the time to testing varied between samples. We suspect that the inhibitory action of haem from the whole blood spotted on DBS cards led to these samples not amplifying. The factors that affect the efficiency of HIV-1 genotyping from DBS are the assay methodology, storage conditions (specifically, humidity and temperature), the length of the storage period, the presence of PCR inhibitors from erythrocytes, and the length of the sequence fragments.^30–32 Bertagnolio et al. state that amplification failure from DBS did not correlate with low plasma viral loads.³³ However, Omooja et al. listed low viral loads as one of the correlates of genotyping failure of DBS samples.² The lowest plasma viral load that amplified from DBS in our study was log₁₀ 2.88 copies/mL. Hence, DBS does show promise, as PLWH in virological failure are expected to have very high HIV-1 viral loads.

Most mutations observed in the study samples were against NRTIs and NNRTIs, the recommended first-line antiretrovirals used during the sample collection period. DRMs described by Zhou et al. showed 90.4% concordance between plasma and DBS, and 9.6% discordance.²⁸ Rottinghaus et al. reported that 93.8% of DRMs detected in plasma were detected from DBS samples.³⁴ Another study found 80% identical drug resistance profiles with plasma and DBS.³⁵ In our study, the observed percentage agreement between plasma and DBS, plasma and PBMCs, and DBS and PBMCs was 71%, 84%, and 74%, respectively. Different factors could lead to the DRM differences observed between the three sample types, including proviral DNA in DBS and PBMCs with mutations distinct from the free virus, inconsistent amplification of viral quasi-species or variants, primer binding preference, sequence quality, and base-calling variability.^28,30 Low-abundance drug-resistant HIV-1 variants (LA-DRVs) or minority variants are present at <20% of the circulating viral population and may not be detected by Sanger sequencing.³⁶ The variability in the distribution of minority variants in plasma, whole blood, and cellular compartments could also impact differences in DRMs. The clinical impact of these variants remains controversial. A systematic review found that 44% of studies reported a significant association between pre-existing LA-DRVs and the risk of virological failure, while 56% showed no association.³⁷ It is important to consider the impact of identifying an additional or reduced number of mutations on clinical interpretation and decision-making.

In addition to plasma and DBS, DRMs can be detected in PBMCs. It is important to note that the emergence of DRMs in plasma precedes that of proviruses in PBMCs by about 425 days.³⁸ The archived mutations will be retained in proviral DNA, even in the absence of drug pressure due to discontinuation or regimen alteration.³⁹ Though the significance of archived mutations in PBMCs is still in question for individuals failing their current regimen, it can be considered complementary to plasma for drug resistance surveillance and an alternative in cases with low plasma viral load levels (100% genotypic success).

The main limitation of our study was the lack of total nucleic acid extraction from DBS, which may have significantly improved the downstream process, amplification, and genotyping success rates. The sample size was limited. Another limitation was that the integrase region was not sequenced from these biosamples. Since INSTIs were added to the first-line ART regimen in India only in 2020, INSTI resistance was not profiled in this study. We had also estimated the HIV-1 viral loads from plasma only, but not from PBMC and DBS specimens, except for troubleshooting the five DBS samples that did not amplify. The storage temperature was not as extreme as in real-life scenarios since the cards were packaged appropriately and left at 25–30°C for 10 days. DBS was spotted with a fixed volume of venous blood using pipettes in a laboratory setting, which may not reflect the actual impact of practical weather conditions, suboptimal DBS preparation, and storage by inexperienced personnel in resource-limited settings. Proper packaging and storage of DBS are vital for successful genotyping.³⁰

This study evaluated the feasibility of using DBS stored at room temperature for HIV-1 genotypic testing and surveillance in resource-limited countries, which could subsequently guide the development of national policies. The use of DBS could help in collecting country-wide representative samples by including specimens from hard-to-reach areas of our country. It expands the options to safely dispatch samples to distant reference laboratories with genomic sequencing capabilities in other countries. The additional role of DBS could be among the children living with HIV, for whom the collection of a large volume of blood is impractical.²⁶ Thus far, the predominant application of DBS has been for surveillance and research; it is not the sample of choice for clinical drug resistance monitoring, unless it is the only feasible sampling option in a particular setting. However, to improve the quality of care given to all PLWH, genotypic drug resistance testing should be performed before any change in the ART regimen.

Conclusion

Although plasma is the ideal sample for detecting DRMs against the current regimen, DBS is a promising alternative for HIV-1 genotypic testing in resource-limited settings due to its ease of sampling, storage, transportation, workforce efficiency, and cost-effectiveness. In addition, PBMCs showed 100% amplification and good correlation with plasma. In conclusion, no single specimen type can satisfy all requirements and purposes. Selecting an appropriate specimen for a setting requires careful consideration of the practical constraints, logistical capacity, and application needs.

Authors’ Contributions

P.S.: Investigation, formal analysis, data curation, writing—original draft, and visualization. D.A.: Data curation, formal analysis, writing—review and editing, and visualization. V.V.R.: Methodology, investigation, and writing—review and editing. J.P.D.: Methodology, investigation, and writing—review and editing. J.F.: Conceptualization, writing—review and editing, visualization, resources. P.A.: Conceptualization, writing—review and editing, and resources. R.K.: Conceptualization, writing—review and editing, resources, supervision, and project administration.

Footnotes

Acknowledgment

This study was conducted as a part of Dr Priyanka Sabu’s M.D. Dissertation under The Tamil Nadu Dr. M.G.R. Medical University, Chennai, Tamil Nadu, India. The authors are grateful to the National AIDS Control Organization (NACO), India, for their continued support, as our laboratory is a part of the nationwide network of HIV reference laboratories (NRL).

Author Disclosure Statement

All authors declare that they have no conflict of interest.

Funding Information

This work was supported by the Fluid Research Grant - IRB Min. No. 9832 dated 07.01.2016, Christian Medical College, Vellore, India, and the Virology Special Fund, Department of Virology, Christian Medical College, Vellore, India.

References

Clipman

, Solomon

, Srikrishnan

, et al. Antiretroviral drug resistance in HIV sequences from people who inject drugs and men who have sex with men across 21 cities in India. Open Forum Infect Dis, 2022; 9(10):ofac481.

Omooja

, Bbosa

, Lule

, et al. HIV-1 drug resistance genotyping success rates and correlates of Dried-blood spots and plasma specimen genotyping failure in a resource-limited setting. BMC Infect Dis, 2022; 22(1):474.

Laboratory Testing: Drug-Resistance Testing | NIH [Internet]. 2024. Available from: https://clinicalinfo.hiv.gov/en/guidelines/hiv-clinical-guidelines-adult-and-adolescent-arv/laboratory-testing-drug-resistance [Last accessed: May 2, 2025].

Human Immunodeficiency Viruses: Collection, Storage and Transport of specimens. ASM Man. 10th edition. Washington DC; pp. 1306–1307.

Holodniy

, Rainen

, Herman

, Yen-Lieberman

. Stability of plasma human immunodeficiency virus load in VACUTAINER PPT plasma preparation tubes during overnight shipment. J Clin Microbiol, 2000; 38(1):323–326.

Hirsch

, Brun-Vézinet

, D’Aquila

, et al. Antiretroviral drug resistance testing in adult HIV-1 infection: Recommendations of an International AIDS Society–USA Panel. JAMA, 2000; 283(18):2417–2426.

David

, Sachithanandham

, Jerobin

, Parasuram

, Kannangai

. Comparison of HIV-1 RNA level estimated with plasma and DBS samples: A pilot study from India (South). Indian J Med Microbiol, 2012; 30(4):403–406.

Masciotra

, Garrido

, Youngpairoj

, et al. High concordance between HIV-1 drug resistance genotypes generated from plasma and dried blood spots in antiretroviral-experienced patients. AIDS, 2007; 21(18):2503–2511.

Johannessen

, Garrido

, Zahonero

, Naman

, de Mendoza

. HIV-1 drug resistance testing from dried blood spots collected in rural Tanzania using the ViroSeq HIV-1 Genotyping System. J Antimicrob Chemother, 2011; 66(2):260–264.

10.

Youngpairoj

, Masciotra

, Garrido

, Zahonero

, de Mendoza

, García-Lerma

. HIV-1 drug resistance genotyping from dried blood spots stored for 1 year at 4°C. J Antimicrob Chemother, 2008; 61(6):1217–1220.

11.

Tran

, Doan

, Phan

HTT

, Doan

, Nguyen

. The effectiveness of HIV drug resistance mutation testing on dried blood spot (DBS) samples in Vietnam. Fortune J Health Sci. Fortune Journals, 2023; 6(4):363–372.

12.

Bertagnolio

, Parkin

, Jordan

, Brooks

, García-Lerma

. Dried blood spots for HIV-1 drug resistance and viral load testing: A review of current knowledge and WHO efforts for global HIV drug resistance surveillance. AIDS Rev, 2010; 12(4):195–208.

13.

Chandane Tak

, Vaidyanathan

, Mukherjee

. Revolutionizing HIV-1 viral load monitoring in India: The potential of dried blood spot analysis for expanding access and improving care. Healthcare (Basel), 2024; 12(4):413.

14.

Raskar

, Thlengliani

, Lalbiaknunga

, et al. Analysis of HIV-1 subtypes and drug resistance mutations in people who inject drugs in Aizawl and nearby districts of Mizoram, India. Virulence. Taylor & Francis, 2025; 16(1):2478074.

15.

Sarmati

, Nicastri

, Uccella

, et al. Drug-associated resistance mutations in plasma and peripheral blood mononuclear cells of human immunodeficiency virus type 1-infected patients for whom highly active antiretroviral therapy is failing. J Clin Microbiol, 2003; 41(4):1760–1762.

16.

, Chen

, Fu

, et al. Consistency of drug-resistant mutations in plasma and peripheral blood mononuclear cells of patients with treatment-naïve and treatment-experienced HIV-1 infection. Front Cell Infect Microbiol, 2023; 13:1249837.

17.

Zhao

, Liao

, Xing

, et al. Evaluation of an in-house HIV-1 drug resistance genotypic testing for using dried blood spot specimens in China. Curr HIV Res, 2015; 13(2):109–116.

18.

Mani

, Ramalingam

, Lionel

, et al. Emergence of HIV-1 drug-resistant variants in women following antiretroviral prophylaxis for the prevention of mother to child transmission. Indian J Med Microbiol, 2015; 33(2):225–230.

19.

Ranjan

, Ponnuvel

, Fletcher

, et al. Evaluation of dried blood spots as a feasible alternative to plasma for the detection and quantification of hepatitis c virus in a tropical setting: A pilot study. Indian J Med Microbiol, 2019; 37(1):60–66.

20.

Kandathil

, Kannangai

, Abraham

, et al. The frequency of HIV-I drug resistance mutations among treatment-naive individuals at a tertiary care centre in south India. Int J STD AIDS, 2009; 20(8):522–526.

21.

Stanford HIV Drug Resistance Database [Internet]. [cited 2025 Mar 29]. Available from: https://hivdb.stanford.edu/

22.

Maartens

, Celum

, Lewin

. HIV infection: Epidemiology, pathogenesis, treatment, and prevention. Lancet, 2014; 384(9939):258–271.

23.

WHO HIV Drug Resistance Report 2021 [Internet]. [cited 2022 Dec 21]. Available from: https://www.who.int/publications-detail-redirect/9789240038608

24.

Singh

, Flores

, Kirby

, et al. Drug resistance in non-B subtype HIV-1: Impact of HIV-1 reverse transcriptase inhibitors. Viruses, 2014; 6(9):3535–3562.

25.

Neogi

, Bontell

, Shet

, et al. Molecular epidemiology of HIV-1 subtypes in India: Origin and evolutionary history of the predominant subtype C. PLoS One, 2012; 7(6):e39819.

26.

, Sandstrom

. Overview of the analytes applied in genotypic HIV drug resistance testing. Pathogens, 2022; 11(7):739.

27.

Diallo

, Murillo

, de Rivera

, et al. Comparison of HIV-1 resistance profiles in plasma RNA versus PBMC DNA in heavily treated patients in Honduras, a resource-limited country. Int J Mol Epidemiol Genet, 2012; 3(1):56–65.

28.

Zhou

, Wagar

, DeVos

, et al. Optimization of a low cost and broadly sensitive genotyping assay for HIV-1 drug resistance surveillance and monitoring in resource-limited settings. PLoS One, 2011; 6(11):e28184.

29.

WHO Manual for HIV Drug Resistance Testing Using Dried Blood Spot Specimens. 3rd ed. Geneva: World Health Organization; 2020.

30.

Parry

, Parkin

, Diallo

, et al. Field study of dried blood spot specimens for HIV-1 drug resistance genotyping. J Clin Microbiol, 2014; 52(8):2868–2875.

31.

Neufeld

, Munyuza

, Reimer

, et al. A validated in-house assay for HIV drug resistance mutation surveillance from dried blood spot specimens. J Virol Methods, 2024; 327:114939.

32.

Rodriguez-Auad

, Rojas-Montes

, Maldonado-Rodriguez

, et al. Use of dried plasma spots for HIV-1 viral load determination and drug resistance genotyping in Mexican patients. Biomed Res Int, 2015;2015;2015:240407.

33.

Bertagnolio

, Soto-Ramirez

, Pilon

, et al. HIV-1 drug resistance surveillance using dried whole blood spots. Antivir Ther, 2007; 12(1):107–113.

34.

Rottinghaus

, Ugbena

, Diallo

, et al. Dried blood spot specimens are a suitable alternative sample type for HIV-1 viral load measurement and drug resistance genotyping in patients receiving first-line antiretroviral therapy. Clin Infect Dis, 2012; 54(8):1187–1195.

35.

Monleau

, Aghokeng

, Eymard-Duvernay

, et al.; ANRS 12235 Study Group. Field evaluation of dried blood spots for routine HIV-1 viral load and drug resistance monitoring in patients receiving antiretroviral therapy in Africa and Asia. J Clin Microbiol, 2014; 52(2):578–586.

36.

Zhou

, Ouyang

, Liu

, et al. A sensitivity and consistency comparison between next-generation sequencing and sanger sequencing in HIV-1 pretreatment drug resistance testing. Viruses, 2024; 16(11):1713.

37.

Mbunkah

, Bertagnolio

, Hamers

, et al.; WHO HIVResNet Working Group. Low-abundance drug-resistant HIV-1 variants in antiretroviral drug-naive individuals: A systematic review of detection methods, prevalence, and clinical impact. J Infect Dis, 2020; 221(10):1584–1597.

38.

, Gatanaga

, Ida

, et al. Emergence of protease inhibitor resistance-associated mutations in plasma HIV-1 precedes that in proviruses of peripheral blood mononuclear cells by more than a year. J Acquir Immune Defic Syndr, 2003; 34(1):1–6.

39.

Banks

, Gholamin

, White

, Zijenah

, Katzenstein

. Comparing peripheral blood mononuclear cell DNA and circulating plasma viral RNA pol genotypes of subtype C HIV-1. J AIDS Clin Res, 2012; 3(2):141–147.