The review included human participants of any age, sex, HIV status, or geographic location who were treated with TFV-based ART for PrEP and treatment. Animal studies and laboratory-based studies without human participants were excluded.

The review included studies measuring TFV-DP concentrations in DBS as a marker of adherence, regardless of study design, or methodology. Eligible studies were those that reported adherence thresholds (number of doses/week) or analysed clinical outcomes in relation to TFV-DP concentrations. Optimal adherence was defined as participants taking all doses prescribed. Adequate adherence was defined as taking the number of prescribed doses needed to achieve clinical protection, such as a reduced risk of HIV acquisition in PrEP studies or virological suppression in treatment studies.

Studies were considered regardless of the healthcare or research setting, including clinical trials, community-based studies, or laboratory investigations. The review focused on studies using quantitative methods, such as pharmacokinetic (PK) analyses, cross-sectional designs, or cohort studies. Narrative reviews, commentaries, and other non-primary literature were excluded.

Ten reference studies were identified (Table 1) which included 2037 participants, mostly from high-income countries. Sample sizes ranged from 12 to 1603 participants. These studies included people without HIV aged 18 to 49 years, and peripartum women with and without HIV. Most reference studies were performed under controlled laboratory or clinical conditions in the United States,^{5,10,11,14,15,16} with some representation from sub-Saharan Africa and Asia (Thailand), extending the evaluation of TFV-DP thresholds to more diverse and resource-limited countries.^6,8,17,18

Four PrEP studies were included (Table 1). Three were prospective cohort studies with sample sizes ranging from 103 to 170 participants,^19,20,21 conducted in the United States, Kenya, and Uganda. The fourth study was a large pooled analysis of 72 PrEP studies, including 17 274 participants across 28 countries, not including the three prospective cohort studies mentioned above.²² Participants were predominantly adults, with median ages ranging from the mid-20s to late-30s, and men (approximately 66% – 100% in earlier cohort studies and 63% in the large pooled analysis).

Eight treatment studies were identified (Table 1) which included 2226 participants, with sample sizes between 36 and 532 participants.^{7,9,23,24,25,26,27} Most studies were conducted in the United States, with three studies conducted in South Africa. Ages ranged from 18 to 52 years and included a higher proportion of women, particularly in peripartum and postpartum studies. Overall, 741 participants were women, representing 33% of the total treatment cohort.

Reference studies used a range of primary study designs, including PK studies, randomised crossover trials, and prospective cohort studies to establish TFV-DP concentrations in DBS based on the number of doses. The data from these studies were used to determine definitions of adherence correlated to TFV-DP concentrations in DBS with different dosing regimens and were cited by subsequent studies in the PrEP and treatment categories (Table 2). The study populations included people without HIV, men who have sex with men (MSM), and peripartum women using PrEP or ART. The duration of the included studies ranged from 1 to 18 months. An adequate adherence threshold for PrEP to prevent HIV transmission was commonly defined as four doses per week, corresponding to a TFV-DP concentration of ≥ 700 fmol/punch in DBS. This threshold was not derived as a pooled average across studies but is based on the most commonly reported concentration from controlled PK and observational studies with directly observed dosing, including Castillo-Mancilla et al.,⁵ Anderson et al.,¹⁶ and Grant et al.,¹⁴ which estimated TFV-DP concentrations associated with specific dosing frequencies. However, it was not consistently observed across studies, and the TFV-DP concentrations associated with similar adherence levels often varied by population and study context. For example, in a PK study conducted in Asia, Niu et al.¹⁷ reported TFV-DP concentrations of approximately 779 fmol/punch associated with adherence equal to or exceeding four doses per week, which highlights variability around the commonly reported ≥ 700 fmol/punch benchmark. TFV-DP concentrations also varied by ART regimen, with lower concentrations observed with efavirenz-based regimens compared with integrase inhibitor-based regimens despite similar adherence and virological suppression (E. Decloedt, personal communication, 2025, unpublished data).

Methodological factors also influenced threshold interpretation. Mugwanya et al.¹⁸ showed that TFV-DP concentrations varied according to DBS extraction methodology, with lower concentrations observed using a 70:30 methanol–water extraction compared with a 50:50 extraction. In this study, adequate adherence (> 4 doses/week) corresponded to ≥ 700 fmol/punch (70:30) and ≥ 900 fmol/punch (50:50), while optimal adherence (7 doses/week) corresponded to ≥ 1250 fmol/punch (70:30) and ≥ 1600 fmol/punch (50:50).

Beyond methodological differences, several population-specific and physiological factors contributed to variability in TFV-DP thresholds across reference studies. For example, studies in peripartum women without HIV defined optimal PrEP adherence as seven doses per week, which corresponds to ≥ 650 fmol/punch during the second to third trimesters of pregnancy, and to ≥ 1050 fmol/punch in the postpartum period.^6,8 Variability was observed in adequate adherence definitions based on body weight, pregnancy status, and TFV formulation. Ibrahim et al.¹¹ introduced a model incorporating weight-based adherence benchmarks, showed that participants weighing ≤ 110 kg required only two to three doses per week to achieve TFV-DP concentrations similar to those weighing > 110 kg taking four doses per week.

The smaller PrEP studies from the United States and Africa included MSM and serodiscordant couples,^19,20,21 whereas the large global pooled analysis included diverse populations²² (Table 3). These studies quantified TFV-DP concentrations in DBS to assess adherence and correlated it with HIV prevention outcomes. While the reference studies established the relationship between dosing frequency and TFV-DP concentrations, PrEP studies reported large variability in observed TFV-DP concentrations, suggesting differences in adherence patterns across populations.

Hoenigl et al.¹⁹ defined adequate adherence as TFV-DP ≥ 719 fmol/punch (≥ 4 doses/week), a threshold derived from the reference study by Grant et al.¹⁴ The observed median TFV-DP concentration was 993 fmol/punch at four doses/week. However, the interquartile range (IQR) varied widely (0–1397 fmol/punch), indicating variability in individual adherence or TFV-DP PK.

Pyra et al.²⁰ used the reference adherence thresholds of ≥ 700 fmol/punch (≥ 4 doses/week) for adequate adherence, as defined by Castillo Mancilla et al.⁵ and the extrapolated concentration of ≥ 1050 fmol/punch (≥ 6 doses/week) for optimal adherence, as defined by Anderson et al.¹⁶ Observed mean TFV-DP concentrations were 925 fmol/punch (standard deviation [s.d.]: 509) (4 doses/week), and 994 fmol/punch (s.d.: 517) (6 doses/week), which demonstrates moderate variability across participants. Among 150 DBS samples from the 103 participants, 87 samples (58%) met the lower adherence cut-off of ≥ 4 doses/week, and 62 (41%) met the ≥ 6 doses/week target.

Wahome et al.²¹ applied the reference adherence threshold of TFV-DP ≥ 700 fmol/punch (≥ 4 doses/week) for adequate adherence, based on Grant et al.’s study,¹⁴ However, the observed TFV-DP concentrations were notably lower than expected, with only 14.5% of participants reaching the adherence threshold. Among participants who acquired HIV while on PrEP, 80% had undetectable TFV-DP concentrations, suggesting poor adherence rather than PrEP failure. The HIV incidence was 3.6 per 100 person-years (PY) in MSM reporting PrEP use, and 5.4 per 100 PY in a concurrently followed internal comparator group who never started PrEP, indicating that suboptimal adherence may have contributed to the lack of risk reduction in this cohort.

In the large pooled analysis of 72 PrEP studies, Landovitz et al.²² evaluated TFV-DP concentrations in DBS as categorical adherence corresponding to dose equivalents. The majority of HIV seroconversions (83.3%) occurred in participants with TFV-DP concentrations consistent with < 2 doses/week, while HIV incidence declined substantially among those with higher adherence, reaching 0.27 per 100 PY in the 4–6 doses/week category, and 0.054 per 100 PY among participants with TFV-DP concentrations consistent with ≥ 7 doses/week.

The eight treatment studies assessed the relationship between TFV-DP concentrations in DBS and virological suppression in PLWH receiving TFV-based ART (Table 4). Across multiple studies, higher TFV-DP concentrations were consistently associated with sustained virological suppression. For example, Castillo-Mancilla et al.²⁵ applied TFV-DP > 1560 fmol/punch as the optimal adherence cut-off (i.e. taking all doses in a week) based on two reference studies.^5,14 For 100% adherence, they reported observed median TFV-DP concentrations in DBS of 1847 fmol/punch (range 706–3776), and a median of 926 fmol/punch (range 706–1133) as adequate adherence in women who largely maintained HIV RNA < 20 copies/mL (n = 8/35), indicating virological suppression despite suboptimal dosing. Castillo-Mancilla et al.⁹ established a strong correlation between TFV-DP concentrations and virological suppression, with a threshold of ≥ 1850 fmol/punch, leading to a 73.5-fold increased likelihood of maintaining HIV RNA < 20 copies/mL versus < 350 fmol/punch. Similarly, Castillo-Mancilla et al.²⁷ further demonstrated that only 21% of virologically suppressed participants (< 20 copies/mL) had TFV-DP ≥ 1850 fmol/punch and detectable emtricitabine triphosphate concentrations, indicating that some PLWH may maintain suppression despite suboptimal adherence. Moreover, Morrow et al.²⁶ found that individuals with TFV-DP < 800 fmol/punch had a significant increase in the risk of developing viraemia. Specifically, compared with TFV-DP ≥ 1650 fmol/punch, TFV-DP < 800 fmol/punch was associated with 4.7-fold higher odds of future viraemia ≥ 20 copies/mL.

In the South African setting, Jennings et al.²³ reported that TFV-DP concentrations < 400 fmol/punch were strongly associated with future viraemia (> 400 copies/mL; odds ratio [OR]: 16.1, 95% confidence interval [CI]: 3.9–67.4). Odayar et al.⁷ studied postpartum women and found that 25% of viraemic events occurred even with TFV-DP ≥ 700 fmol/punch (vs 17% with ≥ 1850 fmol/punch). Importantly, women with TFV-DP < 350 fmol/punch had 12.9-fold higher odds of viraemia (≥ 20 copies/mL) compared to those with ≥ 1850 fmol/punch. Van Heerden et al.²⁸ evaluated TFV-DP concentrations in DBS among adults initiating tenofovir-lamivudine-dolutegravir (TLD) as first- or second-line ART in South Africa. Higher TFV-DP concentrations were associated with virological suppression and lower risk of virological rebound, with TFV-DP concentrations assessed as a continuous variable performing better than categorical thresholds for predicting virological outcomes.

In addition, Robbins et al.²⁴ showed that TFV-DP concentrations in DBS were inversely correlated with HIV viral load, with lower concentrations associated with viraemia (> 200 copies/mL) and higher concentrations linked to virological suppression (< 20 copies/mL). These corresponded with mean TFV-DP concentrations of 825 fmol/punch versus 446 fmol/punch, respectively, for tenofovir disoproxil fumarate (TDF) regimens, and 1847 fmol/punch versus 309 fmol/punch, respectively, for tenofovir alafenamide (TAF) regimens.

In the treatment studies, a TFV-DP concentration of ≥ 1850 fmol/punch was reported as a threshold associated with virological suppression in three out of eight studies.^7,9,27 Other reported benchmarks included a TFV-DP concentration of 1560 fmol/punch as the target for optimal (100%) adherence,²⁵ < 800 fmol/punch as a predictive threshold for future viraemia,²⁶ and TFV-DP ≤ 400 fmol/punch as a strong predictor of viral breakthrough.²³ Although the cut-off points varied across studies, these differences reflect distinct adherence categories and virological outcomes. Collectively, the findings consistently demonstrate that higher TFV-DP concentrations in DBS are associated with sustained virological suppression. Figure 2 summarises the reported TFV-DP thresholds across reference, PrEP, and treatment studies.

Individual dots represent TFV-DP threshold values reported in included studies. Vertical lines represent the range of reported thresholds within each study category. Boxplots represent descriptive summaries (median and interquartile range) of the reported TFV-DP threshold values plotted within each study category. The dashed horizontal lines indicate the commonly reported benchmark for each category. Thresholds represent adherence-, HIV prevention-, virological suppression-, or future viraemia-associated TFV-DP concentrations according to the primary objective of each study.