*2.2. Serum*

Serum is the fluid left after whole blood is naturally clotted. The compositions of plasma and serum are nearly identical except for fibrinogen, which is present in plasma but naturally removed from serum during clotting. Serum is the most commonly used substitute for plasma in molecular HIV assays. While plasma is always the preferred analyte, serum is often used for serological HIV diagnosis. The suitability of serum versus plasma for HIVDRT testing varies largely depending on the availability of centrifugation devices and the usage of anti-coagulant during the sample collection. However, remnant frozen sera from diagnostic testing are often utilized in retrospective HIVDR surveillance and research projects [8].

While applied in HIV molecular assays, another notable difference between plasma and serum analytes is the concentration of HIV contents. A small portion of HIV particles or viral RNA can be trapped in the blood coagulum during clotting, rendering lower VLs in serum than in plasma [11,12]. There is no evidence showing different HIVDR profiles as determined using plasma or serum, implying that the HIV viral RNA loss resulting from clotting is non-selective. However, precautions should still be taken if serum specimens are used for HIVDRT and viral diversity analysis, especially when next generation sequencing (NGS) technologies are used. NGS resolves the intra-host viral diversity and DRMs of lower abundance with a significantly higher resolution than conventional population-based Sanger sequencing [13,14]. Moreover, serum might not be an ideal analyte for patients with lower expected VLs, such as those currently on ART.

The advantages of plasma and serum analytes in HIVDRT are apparent. However, their limitations are also evident. Both analytes require well-trained phlebotomists, skilled lab personnel, stringent cold-chain transportation, and low-temperature storage conditions to maintain the specimen quality and HIV template integrity. These requirements limit the suitability and feasibility of plasma and serum for applications in remote or resourcelimited settings (RLS) [15,16]. Searching for ideal alternative analytes or specimen collection metrics has been an everlasting interest for HIVDR professionals.


**Table 1.** Varied analytes applied in HIV drug resistance test.


**Table 1.** *Cont.*

## *2.3. Whole Blood*

Besides plasma and serum, anticoagulated whole blood (WB) is another commonlyused laboratory analyte. WB is widely used when the isolation of plasma/serum is not feasible, but nucleic acid extraction from WB on time is doable. HIV RNA may retain its integrity for 72 h in WB at an ambient temperature of 25 ◦C [17]. Therefore, WB could be a suitable substitute for HIV molecular assays targeting HIV viral RNA. Despite this, most WB-based HIVDR studies employed only the DNA extracts.

While plasma and serum both contain HIV viral RNA primarily, the HIV-infected cellular components in WB carry HIV proviral DNA. Depending on the nucleic acid extraction strategies applied, the HIV genetic materials recovered from WB could include viral RNA, proviral HIV DNA, or a combination of both if total nucleic acid (TNA) is extracted. Likewise, the application values of the WB specimens vary depending on the HIV templates used in further HIVDRT. Using RNA extract from WB may approximate the results from plasma/serum reflecting the circulating HIV population. In contrast, data from the DNA extracts may convey the information from HIV proviruses, a distinct archival viral population that is not as informative for patient management.

Steegen et al. assessed the feasibility of HIVDRT using DNA extracted from WB and compared it with results from plasma viral RNA [18]. High genotyping success rates were achieved for all specimens with detectable viral loads from plasma viral RNA and DNA from WB. Moreover, HIV protease (PR) and reverse transcriptase (RT) genes were successfully amplified from 67.7% and 61.3% of WB DNA preparations from patients with undetectable plasma VL [18]. While the viral DNA from WB boosts the HIV amplification rates, HIVDR data from such DNA extracts were often discordant with RNA extracts, confirming that they reflect different viral populations [19]. In addition, HIVDR data from DNA extracts showed poor reproducibility, implying a possible founder effect [18]. Furthermore, defective proviruses that harbor stop codons in the HIVDRT target genes are not rare, and excluding such defective proviruses from the whole blood DNA-based HIVDR data would significantly improve its clinical application value [20].

To overcome the limitations of DNA from WB, using RNA or TNA from WB for HIVDRT may be beneficial. Saracino et al. demonstrated that combining viral RNA and DNA in HIVDR typing might help identify more DRMs in the patients and assist in a more informed, effective ART regimen selection [21]. Targeted extraction of the RNA or DNA components from WB by enzymatically removing the other is always an option if fewer confounding data are expected. However, this will inevitably reduce the net HIV application rates.

While WB sampling still requires phlebotomy, this analyte eliminates the need for centrifugation devices unavailable in many decentralized health facilities. Skipping the centrifugation step also reduces professional HIV exposure, artificial errors, or contaminations associated with plasma/serum sample processing. The relative stability of HIV RNA in WB at an ambient temperature also enables centralized lab testing if such specimens could be quickly transferred from the collection site to the testing lab, even in the absence of a cold chain [17].

#### *2.4. Peripheral Blood Mononuclear Cells (PBMCs)*

PBMCs consist of lymphocytes and monocytes isolated from the anti-coagulated whole blood by density gradient centrifugation. DNA extraction is usually performed on PBMCs to recover the cellular DNA containing proviral HIV DNA that is integrated into the HIV-infected cells' genome. The derived DNA can then be used for HIVDRT. The genetic discordance between plasma HIV RNA and proviral HIV DNA from PBMCs has been well-documented [22–25]. Bi et al. showed that plasma viral RNA-based genotyping could detect HIV DRMs up to 425 days earlier than PBMC DNA when the plasma VL was less than 104 copies/mL [26]. It further highlights the slow turnover of the proviral population and the drastic distinction between the HIV proviruses and the circulating viral population in plasma [26]. A higher comparability of data from plasma and PBMCs was achievable only when the HIV duration is ≤2 years, the sample VLs are ≥5000 copies/mL, or when the patient is treatment naïve or off ART [27,28].

Depending on the ultimate HIVDRT objectives, proviral DNA from PBMC may have added value for comprehensive HIVDR profiling when HIV DRMs from HIV proviruses are considered [29]. PBMCs could be an alternative analyte for HIVDRT when using plasma viral RNA is not feasible or unsuccessful [30]. While conventional plasma RNA-based HIVDRT performs poorly on samples of low VLs, proviral DNA can be readily recovered from PBMCs in these patients for an extended period [31–33]. Therefore, PBMC may also satisfy the needs for retrospective HIVDR analysis or population-level surveillance, in which the order of DRM occurrence is less of a consideration.

Interestingly, Armenia et al. showed that, combined with low nadir CD4 counts and a short-term viral suppression history, PBMC-based HIVDR profiling could help predict the potential viral rebound after the ART regimen switch [34]. A bit counterintuitively, one recent study by Peng et al. reported that HIVDR mutants emerges in PBMS DNA months before they could be detected in plasma, suggesting that PBMC DNA could be an effective tool for early HIVDR detection [35]. Notably, this was from studying a single patient infected by HIV-1 CRF01\_AE and experienced multiple ART failure episodes [35]. The validation of these findings in larger studies remains to be conducted. Moreover, Moraka et al. recently showed that HIV DRMs identified in PBMCs are often associated with defective proviral genomes among early-treatment children [36]. It could lead to an overestimated HIVDR profiling if such PBMC DNA-based HIVDR data were applied in patient care.

#### *2.5. Dried Fluid Analytes*

As a more affordable and practical sampling option, dried fluid analytes are increasingly applied in HIVDRT, especially in low- to middle-income countries where the HIV/AIDS pandemic hits the most but optimal sample collection and storage are not always feasible [37]. In such cases, dried fluid specimens collected/dried with different matrices or devices may be collected from peripheral clinics, community sites, or even self-collected from patients' homes and then transferred under natural ambient conditions to laboratories for centralized testing.

HIV genetic materials in such dried analytes remain relatively stable over an extended period under a wide range of ambient temperatures and suboptimal shipping and storage conditions. However, the reduced assay sensitivity, consistency, and reproducibility due to the small sample volume and the inevitable RNA degradation are primary concerns when such dried analytes are applied in HIVDRT. Refining the preparation of such specimens, improving the integrity of the HIV templates they contain, and boosting the analytical sensitivity of such analytes for HIVDRT are all everlasting topics in this field of work.

Several dried fluid analytes that have been applied in HIVDR studies thus far are overviewed below. This list is by no means exhaustive, and more developments in this field should come up in the foreseeable future (Table 1).

#### 2.5.1. Dried Filter Paper Analytes (DFPAs)

DFPAs have been applied in diagnostic tests for decades, mainly due to the low cost and the ease of sample collection, transportation, and storage. The use of filter paper for blood collection dates back to the early 1960s, when dried blood spots (DBS) were first used for phenylketonuria diagnosis in pediatric patients [38]. Since then, filter paper has been used as a collection matrix for different body fluids, and DFPAs have been used for a broad spectrum of laboratory assays.

Depending on the fluids used, DFPAs for HIVDRT consist of conventional DBS, dried plasma spots (DPS), and dried serum spots (DSS). While the integrity of HIV templates in DFPAs inevitably decreases, HIVDRT with such analytes has often been reported, although their performance varies significantly [37,39].
