HIV-1 resistance rarely observed in subjects using darunavir once-daily regimens across clinical studies

Abstract

Background: Darunavir 800 mg once daily (QD) is indicated for HIV-1–infected treatment-naïve and treatment-experienced (without darunavir resistance-associated mutations [RAMs]) individuals, and has been evaluated in phase 2/3 studies with durations between 48 and 192 weeks.

Objective: To summarize the development (or identification) of post-baseline resistance (RAMs and antiretroviral phenotypic susceptibility) among subjects receiving darunavir QD dosing.

Methods: Seven phase 2/3 studies with available genotypes/phenotypes for subjects treated with ritonavir- or cobicistat-boosted darunavir 800 mg QD regimens were assessed: ARTEMIS (NCT00258557; n = 343), GS-US-299-0102 (NCT01565850; n = 153), GS-US-216-0130 (NCT01440569; n = 313), ODIN (NCT00524368; n = 294), INROADS (NCT01199939; n = 54), MONET (NCT00458302; n = 256), and PROTEA (NCT01448707; n = 273). Genotypic analyses were conducted at baseline (except switch studies enrolling virologically suppressed subjects [MONET, PROTEA]). Criteria for post-baseline resistance testing and evaluation of the development (or identification [switch studies]) of RAMs (respective IAS-USA mutations) varied slightly across studies.

Results: Among 1686 subjects treated with darunavir 800 mg QD regimens, 184 had protocol-defined virologic failure; 182 had post-baseline genotypes analyzed. Overall, 4/1686 (0.2%) developed (or had identified [switch studies]) primary protease inhibitor and/or darunavir RAMs (ARTEMIS, n = 1; GS-US-216-0130, n = 1; ODIN, n = 1; MONET, n = 1). Only 1/1686 (<0.1%) subject lost darunavir phenotypic susceptibility (ODIN; possibly related to prior ritonavir-boosted lopinavir virologic failure). Among 1103 subjects using a nucleos(t)ide reverse transcriptase inhibitor (N[t]RTI) backbone, 10 (0.9%) developed ≥ 1 N(t)RTI RAM (8 had the emtricitabine RAM M184I/V).

Conclusions: Darunavir has a high genetic barrier to resistance. Across a diverse population of HIV-1–infected subjects treated with darunavir 800 mg QD regimens, the development of darunavir resistance was rare (<0.1%).

Keywords:

• Darunavir once daily
• Human immunodeficiency virus-1
• Resistance
• Genotypic resistance
• Resistance-associated mutations
• Phenotypic resistance
• Antiretroviral

Introduction

Human immunodeficiency virus (HIV)-1 drug resistance testing can provide important information to guide antiretroviral (ARV) treatment strategy because resistance-associated mutations (RAMs) harbored by the virus may restrict treatment options.^1,2 To minimize the risk of developing RAMs, the use of ARV agents with a high genetic barrier to resistance, such as protease inhibitors (PIs), are desirable.^1,2 In the case of boosted darunavir, potent antiviral activity against wild-type and multidrug-resistant HIV-1 and a high genetic barrier to the development of resistance have been demonstrated.^3–5

Darunavir dosed once daily (QD) has been studied extensively across diverse populations in clinical trials, demonstrating excellent efficacy and safety.^6–12 These studies of varying duration have enrolled both treatment-naïve and treatment-experienced subjects and included a range of combination therapies and monotherapy. Boosted darunavir 800 mg, in combination with 2 nucleos(t)ide reverse transcriptase inhibitors (N[t]RTIs), is a recommended initial ARV treatment option in certain clinical situations for HIV-1 infection per guidelines from the United States Department of Health and Human Services (US DHHS; with ritonavir or cobicistat).¹ Boosted darunavir 800 mg, in combination with 2 N(t)RTIs, is also a recommended initial regimen per the European AIDS Clinical Society (EACS; with ritonavir or cobicistat).² Darunavir QD is indicated for HIV-1–infected individuals who are treatment-naïve or treatment-experienced without darunavir RAMs.¹³ Eleven HIV-1 protease (PR) mutations associated with darunavir resistance were previously identified based on an analysis of a pooled dataset of studies in highly treatment-experienced subjects (subjects had received treatment with darunavir 600 mg and ritonavir 100 mg twice daily).^14,15 Consistent with darunavir having a high genetic barrier to the development of resistance, its virologic efficacy has been shown to be compromised only in the presence of 3 or more darunavir RAMs in the background of a high number (median of 14 or 15, respectively^14,15) of International Antiviral Society–USA (IAS-USA) PI RAMs.¹⁶

The current report is an analysis of the development (or identification) of post-baseline resistance, including primary PI, darunavir, N(t)RTI, and non-nucleoside reverse transcriptase inhibitor (NNRTI) RAMs, as well as loss of ARV phenotypic susceptibility, among HIV-1–infected subjects in the darunavir 800 mg QD dosing arms from seven completed clinical studies at their respective final analysis time points.

Methods

Study designs

This analysis was based on data from seven company-sponsored, phase 2 and 3 studies that investigated darunavir 800 mg QD–containing ARV regimens and had available genotype and phenotype data (ARTEMIS [ClinicalTrials.gov Identifier: NCT00258557], GS-US-299-0102 [NCT01565850], GS-US-216-0130 [NCT01440569], ODIN [NCT00524368], INROADS [NCT01199939], MONET [NCT00458302], and PROTEA [NCT01448707]); detailed study designs have been published for each study.^6–12 Only the darunavir 800 mg QD treatment arms were included in this analysis. As shown in Figure 1, the studies differed in clinical phase, sample size, and final analysis time point. The subject populations also varied in regards to treatment experience (i.e. treatment-naïve or treatment-experienced) and, for treatment-experienced subjects, history of ARV use and viral suppression status (i.e. viral load of ≥50 or <50 HIV-1 RNA copies/mL at screening). Treatment varied in terms of boosting agent (ritonavir or cobicistat) and background regimen.

Figure 1 Study summaries of darunavir 800 mg QD–containing treatment arms

Notes: QD, once daily; DRV, darunavir; HIV-1, human immunodeficiency virus-1; TN, treatment-naïve; rtv, ritonavir; FTC, emtricitabine; TDF, tenofovir disoproxil fumarate; cobi, cobicistat; TAF, tenofovir alafenamide; TE, treatment-experienced; N(t)RTI, nucleos(t)ide reverse transcriptase inhibitor; ETR, etravirine; ARV, antiretroviral. ^aFor each ARV regimen, agents combined in the same tablet are separated by a ‘/’ and those in different tablets are separated by a ‘+’

Three of the seven studies (ARTEMIS, GS-US-216-0130, and ODIN) were pivotal phase 3 studies that evaluated the safety and efficacy of darunavir QD treatment (with N[t]RTIs). ARTEMIS was a 192-week, phase 3 study of treatment-naïve adults; among the 689 enrolled subjects, 343 were included in this analysis (i.e. were treated with darunavir 800 mg QD) and had received ritonavir-boosted darunavir with emtricitabine and tenofovir disoproxil fumarate.⁶ GS-US-299-0102 was a 48-week, phase 2 study of treatment-naïve adults; all 153 subjects were included in this analysis and had received cobicistat-boosted darunavir and emtricitabine, with either tenofovir disoproxil fumarate or tenofovir alafenamide.⁷ GS-US-216-0130 was a 48-week, phase 3b study of both treatment-naïve and treatment-experienced adults; all 313 subjects received cobicistat-boosted darunavir and 2 N(t)RTIs.⁸ ODIN was a 48-week, phase 3 study of treatment-experienced adults; of 590 enrolled subjects, 294 were included in this analysis and had received ritonavir-boosted darunavir and ≥2 N(t)RTIs.⁹ INROADS was a 48-week, phase 2b study of treatment-experienced adults and treatment-naïve adults with transmitted resistance; all 54 subjects were included in this analysis and had received ritonavir-boosted darunavir and etravirine.¹⁰ MONET was a 144-week, phase 3 study of virologically suppressed adults; all 256 enrolled subjects were included in this analysis and had received either ritonavir-boosted darunavir monotherapy or ritonavir-boosted darunavir and 2 N(t)RTIs.¹¹ PROTEA was a 96-week, phase 3b study of virologically suppressed adults; all 273 enrolled subjects were included in this analysis and had received either ritonavir-boosted darunavir monotherapy or ritonavir-boosted darunavir and 2 N(t)RTIs.¹²

Key eligibility criteria related to resistance and viral load at screening varied across studies (Table 1). Viral load inclusion criteria for the studies that enrolled virologically failing subjects ranged from >500 to ≥5000 copies/mL. Studies varied in regards to excluded RAMs and the requirement of genotypic or phenotypic ARV susceptibility. Three studies included eligibility criteria that limited history of previous virologic failures.

Table 1 Key resistance eligibility criteria at screening

Study	Viral load at screening (copies/mL)	Excluded RAMs	Genotypic or phenotypic susceptibility	Virologic failure history
ARTEMIS	≥5000	–	–	–
GS-US-299-0102	≥5000	–	DRV, FTC, TDF^a	–
GS-US-216-0130	≥1000	DRV	DRV, 2 N(t)RTIs^a	–
ODIN	>1000	DRV	DRV, optimized N(t)RTI background^a	–
INROADS	>500	DRV, ETR^b	DRV, ETR^a^,^b	≤2 previous virologic failures while on
				PI-containing ARV therapy
MONET	<50	Primary PI^c	–	No history of virologic failure while on
				previous/current ARV therapy
PROTEA	<50	Primary PI^c	–	No history of virologic failure while on
				previous/current ARV therapy

Notes: RAM, resistance-associated mutation; DRV, darunavir; FTC, emtricitabine; TDF, tenofovir disoproxil fumarate; N(t)RTI, nucleos(t)ide reverse transcriptase inhibitor; ETR, etravirine; PI, protease inhibitor; ARV, antiretroviral; HIV-1, human immunodeficiency virus-1; NNRTI, non-nucleoside reverse transcriptase inhibitor; IAS-USA, International Antiviral Society–USA.

^a For GS-US-299-0102, GS-US-216-0130, and ODIN: susceptibility was based on genotype at screening. For INROADS: susceptibility was based on phenotype at screening.

^b Treatment-naïve subjects who were eligible to enroll in INROADS were required to have HIV-1 with transmitted primary drug resistance that conferred genotypic or phenotypic resistance to either efavirenz or nevirapine; those with multiclass transmitted resistance or with transmitted resistance to only classes other than NNRTIs were not eligible to enroll.

^c IAS-USA primary PI mutations were based on historical genotypes.

Evaluation of resistance

Genotypic and phenotypic analyses were conducted at screening (GS-US-299-0102, GS-US-216-0130) or screening/baseline (ARTEMIS, ODIN, INROADS), except for switch studies that enrolled virologically suppressed subjects (MONET, PROTEA). The criteria for protocol-defined virologic failure (PDVF) varied slightly across studies (Table 2). Resistance testing was performed on samples from subjects experiencing PDVF (except MONET and PROTEA, for which any sample that exceeded the viral load threshold was tested). When feasible, resistance testing was performed at the time of (confirmed [preferable] or unconfirmed) virologic failure and/or at later time points. Post-baseline virologic failure plasma samples were sequenced (population sequencing) and phenotyped at Virco BVBA (Mechelen, Belgium) using the Virco^®TYPE HIV-1 and Antivirogram^® assays, or at Monogram Biosciences (South San Francisco, CA, USA) using the GenoSure^® MG and PhenoSense^® GT (PR-reverse transcriptase [RT]) assays. The viral load cut-off for attempting resistance testing was as low as 50 copies/mL for some studies (ARTEMIS, ODIN, INROADS, and MONET) and 400 copies/mL for other studies (GS-US-299-0102, GS-US-216-0130, and PROTEA; Table 2).

Table 2 Viral load criteria for post-baseline resistance testing

Study	PDVF	VL cut-off for resistance testing
ARTEMIS, ODIN, and INROADS^a	Rebounder: on-study at Week 12 and achieved 2 consecutive VL <50 copies/mL, followed by 2 consecutive VL ≥50 copies/mL or discontinuation with a last observed on-treatment VL ≥50 copies/mL or Suboptimal virologic response: on-study at Week 12 and never achieved 2 consecutive VL <50 copies/mL	≥50 copies/mL
GS-US-299-0102 and GS-US-216-0130	Rebounder: VL <50 copies/mL followed by 2 consecutive VL ≥400 copies/mL, or 2 consecutive >1 log10 increase in VL from nadir or Suboptimal virologic response: VL <1 log10 reduction from baseline at Week 8 visit, confirmed at Week 12	≥400 copies/mL
MONET	2 consecutive VL ≥50 copies/mL	>50 copies/mL (at any time)
PROTEA	2 consecutive VL ≥400 copies/mL	>400 copies/mL (at any time)

Note: PDVF, protocol-defined virologic failure; VL, viral load; TLOVR, time to loss of virologic response.

^a TLOVR (non-virologic failure censored) algorithm.

Evaluation of the development of RAMs (or identification of RAMs, in the case of switch studies) was based on PI, N(t)RTI, and NNRTI IAS-USA mutation lists defined in each study (Figure 2). RAMs were considered to have developed if they were detected post-baseline but not at baseline or screening. Mutations that were identified among subjects with virologic failure in the switch studies could not be compared with screening/baseline sequences to evaluate whether they developed during the study due to being virologically suppressed at screening and, therefore, lacking genotypes. If available, loss of phenotypic susceptibility to ARVs was defined as fold change (FC) in the 50% effective concentration (EC₅₀; in cell-based assays) below or equal to the lower clinical cut-off or biological cut-off at baseline, but above the cut-off value post-baseline.

Figure 2 RAM lists^a

Notes: RAM, resistance-associated mutation; PI, protease inhibitor; DRV, darunavir; N(t)RTI, nucleos(t)ide reverse transcriptase inhibitor; ins, insertion; NNRTI, non-nucleoside reverse transcriptase inhibitor; IAS-USA, International Antiviral Society–USA. ^aMutation lists varied between studies; the IAS-USA 2008 (December) list³⁴ was used in ODIN and MONET, IAS-USA 2009 list³⁵ in ARTEMIS, IAS-USA 2010 list³⁶ in PROTEA, IAS-USA 2011 list³⁷ in INROADS, and adapted IAS-USA lists in GS-US-299-0102 and GS-US-216-0130 (primary PI RAMs: identical to IAS-USA 2009 list except that L33F and I54V were considered primary PI RAMs instead of secondary; N[t]RTI RAMs: IAS-USA 2011 list used in GS-US-299-0102 and a modified list used in GS-US-216-0130). ^bThe L33F mutation was considered a primary PI RAM in GS-US-299-0102, GS-US-216-0130, ODIN, and MONET (considered a secondary PI RAM in ARTEMIS, INROADS, and PROTEA). ^cThe I54V mutation was considered a primary PI RAM in GS-US-299-0102 and GS-US-216-0130 (considered a secondary PI RAM in ARTEMIS, ODIN, INROADS, MONET, and PROTEA). ^dThe N83D mutation was considered a primary PI RAM in INROADS and PROTEA (considered a secondary PI RAM in ARTEMIS, GS-US-299-0102, GS-US-216-0130, ODIN, and MONET)

Results

Resistance at baseline

Across the seven studies, a total of 2328 subjects were enrolled and 1686 of these were treated with darunavir 800 mg QD–based regimens. Among non-virologically suppressed subjects at baseline, the majority in each study harbored HIV-1 subtype B (Table 3). In general, baseline RAMs tended to be more common among treatment-experienced subjects (ODIN and INROADS) than treatment-naïve subjects (ARTEMIS, GS-US-299-0102, and most subjects in GS-US-216-0130). All three studies that enrolled treatment-experienced subjects excluded individuals with darunavir RAMs at screening, but notably, three subjects receiving darunavir QD–based regimens in ODIN had ≥1 darunavir RAM at baseline. For studies of treatment-experienced subjects, previous use of PIs, N(t)RTIs, and NNRTIs was common (Table 4).

Table 3 Baseline virus characteristics

Study	HIV-1 subtype, n (%)		≥1 RAM, n (%)
			Protease		Reverse transcriptase
	B	Non-B	Primary PI	DRV	N(t)RTI	NNRTI
ARTEMIS (n = 342)^a	210 (61)	132 (39)	16 (5)	5 (1)	29 (8)	60 (18)
GS-US-299-0102 (n = 153)	149 (97)	4 (3)	8 (5)	6 (4)	15 (10)	30 (20)
GS-US-216-0130 (n = 313)	301 (96)	12 (4)	10 (3)	0	43 (14)	87 (28)
ODIN (n = 294)	179 (61)	115 (39)	47 (16)	3 (1)^b	219 (75)	210 (71)
INROADS (n = 54)	52 (96)	2 (4)	5 (9)	0	25 (46)	31 (57)

Notes: RAM, resistance-associated mutation; HIV-1, human immunodeficiency virus-1; PI, protease inhibitor; DRV, darunavir; N(t)RTI, nucleos(t)ide reverse transcriptase inhibitor; NNRTI, non-nucleoside reverse transcriptase inhibitor.

^a Data were not available for 1 subject.

^b The presence of DRV RAMs at screening was not allowed for ODIN; however, 3 subjects had ≥1 DRV RAM at baseline.

Table 4 Previous ARV use^a

Study	≥1 PI	≥2 N(t)RTIs	≥1 NNRTI
GS-US-216-0130 (n = 18)	9 (50)	16 (89)	12 (67)
ODIN (n = 294)	159 (54)	291 (99)	258 (88)
INROADS (n = 42)	29 (69)	42 (100)	32 (76)
MONET (n = 256)	191 (75)	256 (100)	146 (57)
PROTEA (n = 273)	219 (80)^b	273 (100)	117 (43)

Notes: ARV, antiretroviral; PI, protease inhibitor; N(t)RTI, nucleos(t)ide reverse transcriptase inhibitor; NNRTI, non-nucleoside reverse transcriptase inhibitor; DRV, darunavir.

^a Data are presented as n (%).

^b One hundred and four (38%) subjects had a history of DRV use.

Baseline phenotypic susceptibility data were available in two studies of treatment-experienced subjects. In ODIN, 100% (291/291) of subjects with available data were susceptible to darunavir; susceptibility to N(t)RTIs in the optimized background regimen was as follows: 18% (53/290) of subjects were susceptible to 1 N(t)RTI, 75% (218/290) to ≥2 N(t)RTIs, and 7% (19/290) were not susceptible to any N(t)RTIs used in their background regimen. In INROADS, 100% (54/54) of subjects were susceptible to both study drugs (darunavir and etravirine).

Development of resistance

Of the 1686 subjects included in this analysis, 184 had PDVF and 182 had post-baseline genotypes analyzed (Table 5). Overall, only 4 of 1686 (0.2%) subjects developed (or, in the case of switch studies, had identified) primary PI and/or darunavir RAMs post-baseline (4 of 184 [2.2%] subjects with PDVF).

Table 5 Development (and identification) of post-baseline RAMs and loss of ARV phenotypic susceptibility across studies

Study	Treatment	Subjects, n	Subjects with PDVF, n (%)	Subjects evaluated for resistance, n	Subjects with ≥1 RAM, n
					Protease	Reverse transcriptase
					Primary PI/DRV	N(t)RTI	NNRTI
ARTEMIS (Week 192)	DRV + rtv + FTC/TDF	343	55 (16)	43	V11I; n = 1^a^,^*	M184I/V; n = 3^b	–
						M184V + K70E; n = 1^b
GS-US-299-0102 (Week 48)	DRV/cobi/FTC/TAF	103	4 (4)	4	0	0	–
	DRV + cobi + FTC/TDF	50	1 (2)	1	0	0	–
GS-US-216-0130 (Week 48)	DRV + cobi + 2 N(t)RTIs	313	15 (5)	15	I84I/V; n = 1^c^,^*	M184V; n = 2^d	–
ODIN (Week 48)	DRV + rtv + ≥2 N(t)RTIs	294	65 (22)	60	V32I + M46I + L76V + I84V; n = 1^e	M184V; n = 1^f	–
						V75I + M184V; n = 1^f
						T215F; n = 1^f^,^*
						T215Y; n = 1^f
INROADS (Week 48)	DRV + rtv + ETR	54	7 (13)	2	0	–	E138K + M230L; n = 1^g
							L100L/I + E138E/G + Y181Y/C; n = 1^g^,^*
MONET^h (Week 144)	DRV monotherapy + rtv	127	21 (17)	31	L33F; n = 1^i^,^*	–	–
	DRV + rtv + 2 N(t)RTIs	129	13 (10)	23	0	0	–
PROTEA^h (Week 96)	DRV monotherapy + rtv	137	1 (1)	2	0	–	–
	DRV + rtv + 2 N(t)RTIs	136	2 (1)	1	0	0	–
Total	1686	184 (11)	182	4 (0.2)^j	10 (0.7)^k	2 (3.7)^l
Pivotal phase 3 studies (ARTEMIS, GS-US-216-0130, and ODIN)^m	950	135 (14)	118	3 (0.3)^j	10 (1.1)^k	–

Notes: Bolded values highlight the n (%) of subjects with ≥1 RAM in the total population. RAM, resistance associated-mutation; ARV, antiretroviral; PDVF, protocol-defined virologic failure; PI, protease inhibitor; DRV, darunavir; N(t)RTI, nucleos(t)ide reverse transcriptase inhibitor; NNRTI, non-nucleoside reverse transcriptase inhibitor; rtv, ritonavir; FTC, emtricitabine; TDF, tenofovir disoproxil fumarate; cobi, cobicistat; TAF, tenofovir alafenamide; ETR, etravirine; QD, once daily; TE, treatment-experienced; TN, treatment-naïve.

* Indicates no loss of phenotypic susceptibility for the corresponding ARVs.

^a Subject developed V11I, a secondary PI RAM, after treatment stop (discontinuation due to noncompliance), with no loss of phenotypic susceptibility to DRV or any PI.

^b For all four subjects, there was a loss of FTC phenotypic susceptibility.

^c One TE subject developed I84I/V, a primary PI and DRV RAM, with no loss of phenotypic susceptibility to DRV or any PI.

^d Both subjects (1 TN and 1 TE) used an N(t)RTI backbone consisting of FTC and TDF; both subjects lost FTC phenotypic susceptibility.

^e Subject developed four primary PI RAMs, three of which were also DRV RAMs (V32I, L76V, and I84V), with loss of phenotypic susceptibility to DRV, amprenavir, atazanavir, indinavir, and nelfinavir (possibly related to a previous virologic failure with rtv-boosted lopinavir treatment); see main text for additional information.

^f The subject who developed M184V used an N(t)RTI backbone consisting of abacavir and lamivudine; this subject lost abacavir and lamivudine phenotypic susceptibility. The subject who developed V75I and M184V used an N(t)RTI backbone consisting of abacavir, lamivudine, and tenofovir disoproxil fumarate; this subject lost abacavir and lamivudine phenotypic susceptibility. The subject who developed T215F used an N(t)RTI backbone consisting of lamivudine, tenofovir disoproxil fumarate, and zidovudine; this subject did not lose phenotypic susceptibility to any of the background ARV agents. The subject who developed T215Y used an N(t)RTI backbone consisting of lamivudine, tenofovir disoproxil fumarate, and zidovudine; this subject lost zidovudine phenotypic susceptibility.

^g The subject who developed E138K and M230L lost ETR etravirine phenotypic susceptibility. The subject who developed L100L/I, E138E/G, and Y181Y/C had reduced susceptibility to ETR at the endpoint.

^h RAMs that were identified during the 2 switch studies (MONET and PROTEA) are reported. One MONET subject (DRV + rtv +2 N[t]RTIs arm) who had detectable primary PI and N(t)RTI RAMs was excluded because these mutations were already present at the start of the study as a result of a previous virologic failure.

ⁱ Subject had identified L33F, a primary PI and DRV RAM, which did not confer phenotypic resistance to DRV; see main text for additional information.

^j Percentages are reported in parentheses; the denominators are the total number of treated subjects (the denominator for the “Total” row is 1686 subjects, and the denominator for the “Pivotal phase 3 studies” row is 950 subjects).

^k Percentages are reported in parentheses; the denominators are the total number of subjects who used an N(t)RTI backbone (the denominator for the “Total” row is 1368 subjects, and the denominator for the “Pivotal phase 3 studies” row is 950 subjects).

^l The percentage is reported in parentheses; the denominator is the total number of subjects who used an NNRTI (54 subjects).

^m Combined data from the three pivotal phase 3 studies of DRV QD.

Among the four subjects, one (ARTEMIS) developed V11I, a darunavir (and secondary PI) RAM, after treatment was discontinued due to noncompliance; no loss of phenotypic susceptibility to darunavir or any PI was observed.

A second subject (treatment-experienced, GS-US-216-0130) developed I84I/V, a primary PI and darunavir RAM; this subject did not lose phenotypic susceptibility to darunavir or any PI.

A third subject (ODIN), at Week 48, developed M46I, a primary PI RAM, as well as V32I, L76V, and I84V, all of which were primary PI and darunavir RAMs. This subject was a nonresponder (on-study at Week 12 and never achieved 2 consecutive viral loads <50 copies/mL) who did achieve a viral load of <50 copies/mL at 1 visit (Week 12). The subject lost phenotypic susceptibility to darunavir (FC of 24.4), amprenavir, atazanavir, indinavir, and nelfinavir, which was possibly related to a previous virologic failure with ritonavir-boosted lopinavir treatment (September 2005–March 2008) because all of the developing primary PI RAMs (listed above) and most of the developing secondary PI RAMs (L10I, K20R, I54V, and A71V) are also mutations associated with lopinavir resistance.¹⁶ The subject had also failed an efavirenz-based regimen after the lopinavir treatment failure. At the baseline visit for ODIN, the subject already had 2 PI RAMs (the lopinavir RAMs L63P and V77I) and 4 N(t)RTI RAMs (M41L, M184V, L210W, and T215Y); however, ultra-deep sequencing (at 1% sensitivity, as previously described¹⁷) had not detected the developing mutations at the baseline visit.

A fourth subject (ritonavir-boosted darunavir monotherapy treatment arm of MONET) had L33F, a primary PI and DRV RAM, detected at Week 12 with a viral load of 63 copies/mL, but was resuppressed at subsequent visits until Week 144. The darunavir RAM did not confer phenotypic resistance to darunavir; moreover, it is possible that the mutation was already present at baseline (no sequence available due to suppressed viral load).

Among subjects treated with ritonavir-boosted darunavir monotherapy, only 1 of 264 (0.4%) had a post-baseline darunavir RAM. Overall, among all 1686 subjects treated with any regimen containing darunavir QD included in the analysis, only 1 (<0.1%) lost phenotypic susceptibility to darunavir.

Ten of the 1103 (0.9%) non-virologically suppressed subjects who used an N(t)RTI backbone developed N(t)RTI RAMs; 9 of these subjects also lost phenotypic susceptibility to ≥1 of the N(t)RTIs included in their regimen (Table 5). Four subjects from ARTEMIS, all of whom used emtricitabine and tenofovir disoproxil fumarate, developed RAMs (M184I/V, n = 3; M184V and K70E, n = 1) and lost emtricitabine phenotypic susceptibility. Two subjects from GS-US-216-0130, both of whom used emtricitabine and tenofovir disoproxil fumarate (1 treatment-naïve and 1 treatment-experienced), developed M184V and lost emtricitabine phenotypic susceptibility. Four subjects from ODIN developed N(t)RTI RAMs: 1 subject (who used abacavir and lamivudine) developed M184V and lost abacavir and lamivudine phenotypic susceptibility; 1 subject (who used abacavir, lamivudine, and tenofovir disoproxil fumarate) developed V75I and M184V and lost abacavir and lamivudine phenotypic susceptibility; 1 subject (who used lamivudine, tenofovir disoproxil fumarate, and zidovudine) developed T215F without loss of phenotypic susceptibility to any of the background ARV agents; and 1 subject (who used lamivudine, tenofovir disoproxil fumarate, and zidovudine) developed T215Y and lost zidovudine phenotypic susceptibility.

Among the 54 subjects in INROADS who were treated with ritonavir-boosted darunavir and etravirine, 2 developed NNRTI RAMs (Table 5). One subject developed E138K and M230L and lost phenotypic susceptibility to etravirine. The other subject developed L100L/I, E138E/G, and Y181Y/C, and had reduced susceptibility to etravirine at the endpoint (FC of 8.1).

Discussion

Across a diverse population of HIV-1–infected subjects taking a darunavir 800 mg QD–based regimen, only 4 of 1686 (0.2%) subjects developed (or had identified) post-baseline primary PI and/or darunavir RAMs, and only 1 subject lost darunavir phenotypic susceptibility. The extremely low prevalence of the development of darunavir RAMs and darunavir phenotypic resistance is consistent with the known high genetic barrier to resistance of darunavir. Notably, among treatment-naïve subjects treated with darunavir QD, none developed a primary PI RAM or lost phenotypic susceptibility to darunavir. Among those subjects who were treatment-experienced but PI-naïve, none developed phenotypic darunavir resistance. Importantly, the 1 treatment-experienced subject who lost darunavir phenotypic susceptibility had a prior history of virologic failure with ritonavir-boosted lopinavir treatment, which may have contributed cross-resistant PI RAMs.^14,15 The fact that many of this subject’s developing mutations were lopinavir RAMs suggests the possibility that archived resistance variants may have rapidly reemerged upon exposure to darunavir.

The development of resistance to N(t)RTIs in subjects’ backbone therapies was also rare in these studies. Ten of 1103 (0.9%) non-virologically suppressed subjects who used an N(t)RTI backbone developed N(t)RTI RAMs; the most common, the emtricitabine RAM M184I/V, was found in 8 subjects. Nine of these 10 subjects lost phenotypic susceptibility to ≥1 N(t)RTI in their background regimen.

Boosted-darunavir monotherapy is not recommended in guidelines from the US DHHS, and is an option only for exceptional persons who are not candidates for dual therapies per EACS guidelines, due to reduced efficacy compared with combination regimens^1,2; however, numerous studies have evaluated monotherapy as an ARV regimen simplification strategy in virologically suppressed individuals. Such studies allow for insights into the true genetic barrier to the development of resistance when subjects are treated with only boosted darunavir 800 mg QD. In the current analysis, among those receiving ritonavir-boosted darunavir monotherapy (in MONET and PROTEA), only 1 of 264 (0.4%) subjects was observed to have a darunavir RAM, (which, by itself, does not confer phenotypic resistance to darunavir).^14,15 Across six additional studies of ritonavir-boosted darunavir monotherapy (818 subjects in total), no post-baseline darunavir RAMs were detected and only 2 (0.2%) subjects had a post-baseline primary PI RAM.^18–24 Taken together, findings from studies of boosted-darunavir monotherapy further illustrate its high genetic barrier to resistance. This is in contrast to integrase inhibitor monotherapy, which has resulted in high rates of resistance development.^25,26

The rare development of darunavir resistance upon virologic failure, as seen in the current analysis, is consistent with findings based on a database analysis of US clinical samples.^5,27 Samples (from patients receiving varying ARV/PI regimens and with different treatment experiences and histories of resistance [no prior treatment information was available]) had been submitted for resistance testing (combined genotypic and phenotypic resistance testing) to Monogram Biosciences and were collected from 2006 through 2015.^5,27 The overall prevalence of darunavir RAMs among these submitted samples decreased over time; a corresponding decrease in darunavir phenotypic resistance was also observed.^5,27 The prevalence of darunavir phenotypic resistance remained lower than that of other PIs for each year during the analysis period.^5,27

Of note, there is another darunavir 800 mg QD–based regimen that is currently in clinical development. It is the first single-tablet regimen containing darunavir and combines darunavir 800 mg, cobicistat 150 mg, emtricitabine 200 mg, and tenofovir alafenamide 10 mg. This agent was evaluated in GS-US-299-0102 through 48 weeks⁷ (included in the current analysis and showing no resistance to darunavir, emtricitabine, or tenofovir alafenamide) and is being evaluated in 2 ongoing phase 3 studies (AMBER [NCT02431247] and EMERALD [NCT02269917]²⁸); it was approved in Europe in September 2017 and submitted for regulatory approval in the United States in 2017. This potential regimen includes tenofovir alafenamide, a novel prodrug of tenofovir that provides comparable efficacy at one-tenth of the dose of tenofovir disoproxil fumarate, resulting in a ~90% lower tenofovir plasma concentration and fewer adverse effects, particularly renal and bone adverse effects.^7,29,30 Moreover, strategies to simplify ARV regimens and decrease pill burden, such as single-tablet regimens, may help improve treatment adherence, which may in turn lead to higher virologic suppression rates and further reduce the potential for development of drug resistance.^31–33

In summary, HIV-1 resistance to darunavir was rarely observed in the seven clinical studies of darunavir 800 mg QD–based treatment regimens that were included in this analysis; only 1 of 1686 (<0.1%) subjects lost darunavir phenotypic susceptibility. These findings are consistent with those from previous studies^3–5 and demonstrate that darunavir dosed QD, which is indicated in treatment-naïve and treatment-experienced (without darunavir RAMs) patients, has a very high genetic barrier to the development of resistance.

Funding

This work was supported by Janssen Scientific Affairs, LLC. E.L. and S.D.M. are employees of Janssen Infectious Diseases BVBA. E.Y.W., D.L., S.S., and K.B. are employees of Janssen Scientific Affairs, LLC.

Contributors

E.L., E.Y.W., D.L., S.S., S.D.M., and K.B. contributed to the conception and design of the analysis, interpretation of the data, and drafting and revision of the article.

Acknowledgments

The authors would like to thank Anne Ghys of Janssen Infectious Diseases BVBA for her contributions. Medical writing support was provided by Courtney St. Amour, PhD, of MedErgy, and was funded by Janssen Scientific Affairs, LLC.