Table 4.
Studies reporting pharmacokinetic outcomes with co-administration of hormonal contraceptives and antiretrovirals.
| Reference; location | Design objective (s) | Number of participants (N); Population | Intervention/treatment | Results | Strengths, weaknesses; funding |
| Scarsi et al. [35]Uganda | Prospective open-label pharmacokinetic studyTo assess the effect of EFV-containing or NVP-containing cART on pharmacokinetics of LNG implants in women living with HIV | Sixty HIV+ women >18 years (median age 31), medically eligible for LNG implant, not recently pregnant, and not using other potentially interacting drugs | LNG implantcART-containing NVP or EFV or no cART | Primary endpoint: LNG levels at 24 weeks (compared with no cART group) EFV group ↓47% NVP group ↑35%Secondary: LNG levels at 48 weeks EFV group ↓57% NVP group ↑14%EFV and NVP levels not affected by LNG | Strengths: clearly described interventions and outcomes; valid assays; HIV+ womenWeaknesses: nonrandomized; open label; LNG levels much higher than previously seen in other studiesFunded by government |
| Song et al. [42]USA | Randomized; double-blind; placebo-controlled; crossover studyTo examine the effect of DTG on pharmacokinetic and pharmacodynamic of EE and NGMN in COC users | Sixteen healthy women; age 18–40; mean age 31 years; 94% white; BMI 19–30 (mean 24.7); normal liver function; women had to use a second nonhormonal contraceptive | COC-containing NGMDTG 50 mg or placebo twice daily for 11 days | EE levels unchanged NGMN AUC; Cmin; Cmax unchangedDTG levels similar to historical controls | Strengths: randomized; clearly described methods; valid assaysWeaknesses: healthy women only; single cycleFunded by industry |
| Luque et al. [39]USA | Open-label nonrandomized pharmacokinetic studyTo assess the effect of LPV/r on DMPA pharmacokinetics and vice versa; and to assess safety and tolerability | Twenty-four nonpregnant premenopausal HIV+ women with no recent DMPA; median BMI 28; HIV RNA <400 copies/ml | DMPA; LPV/r containing cARTcART-containing LPV/r | MPA levels compared with historical controls: AUC↑ 46% Cmax ↑66%No changes in LPV or RTV levels after DMPA | Strengths: clearly described population and methods; valid assays; HIV+ womenWeaknesses: small sample size; historical controls; levels only assessed through 12 weeksFunded by government |
| Todd et al. [59]Kenya | Secondary analysis of PrEP HIV prevention trialTo examine pharmacokinetic of LNG with concurrent use of TDF/FTC as PrEP | Twenty-nine healthy women who elected to received LNG implant; ages 18–35TDF/FTC group: N = 17Placebo group: N = 12Mean BMI 22.6 | LNG implantTDF/FTC or placebo | Follow-up 36 weeksLNG levels all above 400 pg/ml; mean LNG levels 469–660 pg/mlLNG levels lower for women randomized to the TDF/FTC arm, but in multivariable analysis TDF/FTC use not associated with changes in LNG levels compared with placebo | Strengths: TDF levels measured to assess for adherenceWeaknesses: Small sample size; percentage retention not stated; LNG concentration measures were missing for 42 time pointsFunded by government |
| Landolt et al. [15,19]Thailand | Open-label; nonrandomized clinical trialTo evaluate EE and ENG levels in women taking EFV-containing and NVP-containing cART and COCs | Forty-eight HIV+ nonpregnant, nonlactating women; 18–45 years, with regular menses, on EFV-containing or NVP-containing cART; no smoking, recent injectable contraceptive use, or contraindications to COCsForteen HIV− controls | COC-containing DSG for two cyclesNVP-containing or EFV-containing cART or no cART | NVP group: EE Cmin ↓58% ENG Cmin↓22%EFV group: EE Cmin ↓9% ENG Cmin↓61%One woman (6%) had NVP level below therapeutic level of 3.1 mg/lThree women had EFV level below therapeutic level of 1.0 mg/l | Strengths: prospective clinical trial; COCs and antiretrovirals at steady state; HIV+ womenWeaknesses: nonrandomized; small sample size; single measurement of ENG levels; high drop out/loss to follow-up rate; unable to measure ENG in 8/16 due to assay interferenceFunded by government |
| Vieira 2014 [34]Brazil | Prospective cohort with pharmacokinetic analysisTo evaluate the effects of EFV-containing or LPV/r-containing cART on ENG implant pharmacokinetics and to determine the impact of cART on luteal activity | Forty-five HIV+ women with regular menstrual cycles, BMI 18–30; excluded women with recent pregnancy or hormonal contraceptive use, acute infections or other opportunistic illnesses, drug or alcohol addiction, use of other potentially interacting drug, chronic diarrhea or malabsorption or noncompliance with cART | ENG implantLPV/r-containing or EFV-containing cART or no cART | ENG levels through 24 weeksEFV group: ENG AUC↓63% Cmax↓54% Cmin ↓70%LPV/r group: ENG AUC↑52% Cmax ↑61% Cmin ↑ 34% | Strengths: Clearly described population and methods; valid assays, HIV+Weaknesses: Small sample size; nonrandomizedFunded by government |
| Atrio et al. [16]Atrio et al. [56]Dubois et al. [17]USA | Open-label; nonrandomized; clinical trialTo compare NET pharmacokinetic in women taking cART with PIs compared with women receiving other cART regimens | Thirty-five HIV+ women age 18–44 years; no changes in medications; no recent hormonal contraceptives; no immunocompromise; no liver or renal disease; normal ovulation; BMI <40; >30 days postpartum; | POPs containing NETIn PI-containing cART group: 11 taking ATV (10/11 on ATV/r); 3 DRV/r; 2 LPV/rIn control group: four women not taking PI-containing cART; 13 taking combinations including ETR, RPV, TDF, FTC, and RAL | PI group: NET AUC ↑50% NET Cmax ↑33% NET Cmin ↑26%In subanalysis limited to women on ATV/r;NETAUC ↑35% NET Cmin ↑39% NET C24 ↑67% | Strengths: clearly described population and methods; valid assays; HIV+ womenWeaknesses: nonrandomized; small sample size; cART use self-reportedFunded by government |
| Crauwels et al. [27]UK | Open-label, three period pharmacokinetic studyTo evaluate the effect of RPV on COC pharmacokinetics and vice versa, and assess effects on sex hormones and safety of co-administration | Eighteen healthy nonsmoking women, 18–45 years; BMI 18–30 (median 24.6); 67% white; excluded pregnant, breast-feeding, or menopausal women, those with history of drug/alcohol abuse, skin disease, or any significant medical problems; use of concomitant medication | COC-containing NETIn third cycle; RPV 25 mg daily days 1–15 | Thirteen completed trialEE Cmin and AUC unchanged EE Cmax ↑17%NET AUC; Cmin; Cmax unchangedRPV pharmacokinetic unchanged from historical controls | Strengths: clearly described population and methods; valid assays; directly observed therapyWeaknesses: healthy women; short-term dosing; single antiretroviral; high discontinuation rateFunded by industry |
| Carten et al. [52]USA | Open-label two period pharmacokinetic studyTo determine the effect of EFV on the pharmacokinetics of LNG EC and vice versa, and assess safety | Twenty-four healthy women; 18–45; normal BMI (mean BMI 27) with no recent use of hormonal contraceptives or other interacting medications; women were either sterilized or used two nonhormonal contraceptive methods | LNG ECPs (0.75 mg) at 0 and 12 h on days 0 and 17EFV 600 mg 72 h after day 0, for 14 days. | Twenty-one women completed studyLNG AUC ↓58% LNG Cmax ↓45% LNG Cmin ↓69% | Strengths: clearly described population and methods; valid assaysWeaknesses: small sample size; healthy women; single antiretroviralFunded by industry |
| Piscitelli et al. [29]UK | Randomized crossover pharmacokinetic studyTo examine if GSK2248761 (fosdevirine) interacts with CYP450 substrates, including COCs | Ten healthy women, without hepatitis and not taking any medications | COC-containing DRSPGSK228761 200 mg or placebo days 1–11 | EE levels unchangedDRSP AUC, Cmax, Cmin ↑18–22% | Strengths: randomizedWeaknesses: short term administration; healthy women; single antiretroviral; small sample size; very few study details provided; trials of fosdevirine on hold due to other safety concerns; study terminated early for unknown reasonsFunded by industry |
| Kasserra et al. [57]USA | Randomized crossover pharmacokinetic studyTo determine the effect of vicriviroc alone or with RTV on COC pharmacokinetics and evaluate safety | Twenty-seven healthy nonpregnant women, 18–40 years; no recent medication use other than acetaminophen or oral contraceptives; BMI 19–32 (median 24.5) | COC-containing NETVicriviroc 75 mg BID for 10 days then vicriviroc 30 mg plus RTV 100 mg daily for next 11 days; group 2: RTV 100 mg daily for first 10 days then plus vicriviroc 30 mg daily for next 11 days | VCV alone: EE AUC and Cmax unchanged NET AUC and Cmax unchangedRTV alone: EE Cmax ↓11% EE AUC ↓29% NET Cmax ↓11% NET AUC ↓7%VCV+RTV: EE Cmax ↓24%% EEAUC ↓29% NET: Cmax ↓11% NET AUC ↓17%No severe or serious AEs | Strengths: randomized; valid assaysWeaknesses: healthy women; small sample size; single antiretroviralFunded by industry |
| Anderson et al. [61]USA | Randomized crossover pharmacokinetic studyTo assess the effect of RAL on COC pharmacokinetic | Twenty healthy nonobese nonpregnant women 18–45 years (mean age 27) using additional barrier contraceptive | COC-containing NGMRaltegravir 400 mg twice daily or placebo days 1–21 | Nineteen women completed the trialEE levels unchangedNGMN AUC ↑ 14% NGMN Cmax ↑29%.No serious clinical or laboratory AEs | Strengths: Randomized; placebo-controlled; clearly described population and methods; valid assaysWeaknesses: evening dose of RAL on day 21 of both periods missed; small sample size; healthy women; single antiretroviralFunded by industry |
| Zhang et al. [55]USA | Open-label three period pharmacokinetic studyTo assess the impact of RTV-boosted ATV on COC pharmacokinetic | Twenty healthy nonpregnant nonbreast-feeding women, 18–45 years (mean age 28); BMI 18–32 (mean 25) | COC containing NGMIn third cycle; ATV/r 300 mg/100 mg daily days 1–14 | EE AUC ↓19% Cmax ↓16% Cmin ↓37%NGMN Cmax ↑ 68% AUC ↑85% Cmin ↑102%Dose normalization estimate magnitude of reduction with lower dose EEATV AUC ↑20% than historical controlsMore AEs with co-administration than with COCs alone (vomiting; headache and abdominal pain); no deaths or SAEs | Strengths: Clearly described population and methods; valid assaysWeaknesses: small sample size; healthy women; single antiretroviralFunded by industry |
| Stuart et al. [30]Malawi | Open label; nonrandomized clinical trialTo assess the feasibility of measuring anovulation in a pharmacokinetic study of COCs and antiretrovirals | Nine women ages 21–35 (3/group) with similar age and BMI: group 1 included HIV+ women on cART; group 2 included HIV+ women not on cART; and group 3 included HIV− women | COC with NGNVP-containing cART or no cART | LNG and EE levels measured by radioimmunoassayLNG AUC 147 in NVP group; 114 no cART group; and 38 in HIV− womenEE AUC; 1384, 1457, 1144, respectivelyAntiretroviral pharmacokinetic similar to historical controls | Strengths: Clearly described population and methods; valid assays; included HIV+ womenWeaknesses: small sample size; nonrandomized; progesterone measured only 1 dayFunding source not reported |
| Sevinsky et al. [26]USA | Open-label three period pharmacokinetic studyTo examine effect of EFV on pharmacokinetics of EE and NGMN and vice versa | Twenty-eight healthy women; 18–45 years (median 26); BMI 20–32 (median 25); on COCs for at least 2 months and no baseline safety issues or breakthrough bleeding | COC-containing NGMEFV 600 mg daily for 14 days during third cycle | Nineteen women completed studyEE AUC, Cmax, Cmin unchangedNGMN Cmax ↓46%; AUC ↓64% Cmin ↓82%Posthoc LNG Cmax; AUC; and Cmin ↓80–86%EFV levels comparable to historical controls | Strengths: clearly described population and methods; valid assaysWeaknesses: small sample size; healthy women; single antiretroviral; high discontinuation rateFunded by industry |
| Vogler et al. [37]USA | Open-label; nonrandomized; clinical trialTo evaluate pharmacokinetic interactions between LPV/r and contraceptive patch and COCs | Thirty-two nonpregnant; premenopausal HIV+ women >13 years either on stable LPV/r regimens for at least 14 days (n = 8) or not on cART or taking NRTIs only (n = 24); nonsmoking; median weight 72 kg; no recent use of injectables or COCs; <198 lb; not taking enzyme inducers | EE/NGMN contraceptive patchSingle-dose COC-containing NETLPV/r (400 mg/100 mg twice a day) or no cART/NRTIs only | Patch EE AUC ↓45% Cmin ↓25%NGMN AUC ↑83% Cmin ↑134%COC EE AUC ↓55%LPV AUC ↓19% Cmin ↓27% Cmax ↓22%RTV AUC↓24% Cmin ↓14% Cmax ↓8% | Strengths: clearly described population and methods; valid assays; HIV infectedWeaknesses: study closed prematurely due to slow accrual; small sample size; single dose COC; high loss to follow-up; not randomizedFunded by government |
| Schöller-Gyüre et al. [28]USA | Open-label three period pharmacokinetic studyTo assess the effect of ETR on COC pharmacokinetics | Twenty-four healthy women; 18–45 years (median 24); BMI 18–30; 97% white; nonsmoking; not pregnant or breast-feeding; no contraindications to hormonal contraceptives; not taking enzyme inducers | COC containing NETIn cycle 3; 200 mg ETR twice daily from day 1 to 15 | EE AUC ↑ 22% Cmax ↑33% Cmin ↑9%NET Cmax; AUC unchanged Cmin ↓22%ETR levels higher than historical controls | Strengths: Clearly described population and methods; valid assaysWeaknesses: small sample size; healthy women; single antiretroviral; not randomizedFunded by industry |
| Kearney and Mathias [58]USA | Open-label; two period pharmacokinetic studyTo evaluate the effect of TDF on the pharmacokinetic of COCs | Twenty healthy nonpregnant; nonlactating women; 19–45 years taking study COC for at least 3 months; without recent medication use or active alcohol or drug use; mean age 25 years; mean weight 64 kg | COC NGMTDF 300 mg/day on days 15–21 of contraceptive cycle 2 | No change in NGMN and EE pharmacokineticTFV levels similar to historical dataNo serious AEs and no discontinuations due to AEs; AEs reported by 10 participants; most commonly headache; rash; dysmenorrhea; nausea; and rhinitis AEs reported by 10 participants; most commonly headache; rash; dysmenorrhea; nausea; and rhinitis | Strengths: clearly described population and methods; valid assaysWeaknesses: healthy women; single antiretroviral; small sample sizeFunded by industry |
| Abel et al. [60]USA | Randomized controlled; crossover pharmacokinetic studyTo evaluate the effect of MVC on COC pharmacokinetic and to assess pharmacokinetic and safety of MVC in women | Fifteen healthy sterilized white women; age 32–45; between 7 and 76 kg (mean 65 kg) and BMI 18–30 | COC LNGMVC 100 mg twice daily or placebo days 1–10 and am of day 11 | EE AUC; Cmax unchangedLNG AUC; Cmax unchangedMVC pharmacokinetic within the range seen in healthy males in previous studiesNo clinically significant abnormalities or severe or serious AEs | Strengths: randomized; placebo controlled; clearly described population and methods; valid assays; 100% retentionWeaknesses: small sample size; healthy women; single antiretroviral; COC only given days 2–8Funded by industry |
| Nanda et al. [32]Brazil | Open-label; nonrandomized; pharmacokinetic studyTo evaluate the effect of EFV-containing cART on pharmacokinetics of MPA and to determine effects on suppression of ovulation and bleeding patterns | Thirty-three HIV+ women aged 19–40 years; with regular menstrual cycles and BMI 18–30 kg/m2; not recently pregnant or breast-feeding | DMPA 150 mg given at enrollmentEFV-containing cART | No difference in MPA levels between groups through 12 weeks | Strengths: Clearly described population and methods; valid assays; HIV+ womenWeaknesses: small sample size; no EFV levels; no MPA levels beyond 12 weeksFunded by government |
| Sekar et al. [38]Belgium | Randomized crossover pharmacokinetic studyTo investigate the effect of DRV/r on COC pharmacokinetics and to examine safety | Twenty-two nonsmoking healthy women, 18–45 years; BMI 18–30; using a second nonhormonal contraceptive method | COC containing NETDRV/r 600/100 mg twice daily days 1–14 or no treatment | Eleven women completed studyEE AUC ↓44% Cmin ↓62% Cmax ↓32%NET AUC ↓14% Cmin ↓30% Cmax ↓10%DRV and RTV levels comparable to historical controls | Strengths: clearly described population and methods; valid assays; randomizedWeaknesses: small sample size; healthy women; single antiretroviral; high discontinuation rateFunded by industry |
| Cohn et al., Watts et al. [14,31]USA | Nonrandomized open-label pharmacokinetic studyTo evaluate the effect of various antiretrovirals on pharmacokinetics of MPA and vice versa; and to determine effects on suppression of ovulation and adverse events | Seventy-two HIV+ nonpregnant; premenopausal women, 22–46 years (median 35); median weight 71 kg; with no recent potentially interacting drugs; women required to use second nonhormonal method of contraception | DMPAAntiretrovirals regimens containing NFV; EFV; NVP; or no antiretroviral/NRTI only | No difference in MPA AUC; Cmax; Cmin; clearance half-life between groupsNVP AUC slightly higher after DMPA. No changes in EFV; nelfinavir pharmacokinetic | Strengths: clearly described population and methods; valid assays; HIV+ womenWeaknesses: nonrandomized; small sample sizeFunded by government |
| Aweeka et al. [65]USA | Open-label two-period pharmacokinetic time series studyTo investigate the effects of sex and contraceptives on ZDV pharmacokinetic and HIV viral load; to evaluate the effect of COCs and DMPA on plasma and genital HIV load among women on stable cART | HIV+: 18 men and 20 women; 22–52 years; on stable ZDV-containing cART | COC containing NET or DMPA begun at second cycle and continued through third cycleParticipants randomized to oral or intravenous ZDV (200 mg)Antiretrovirals included indinavir; nelfinavir; or any NRTI except d4T or tenofovir | Fourteen women (eight DMPA; six COC) provided pharmacokinetic dataNo effect on plasma or cervical HIV viral loadZDV levels and levels of by radioimmunoassay and liquid chromatography/mass spectrometryNo change in ZDV pharmacokinetic after contraceptive useNo differences between COC and DMPA on ZDV pharmacokinetic | Strengths: HIV+; prospective pharmacokinetic study; both PO and IV dosing of ZDVWeaknesses: open-label; nonrandomized; small sample; study stopped early due to slow enrollment; ZDV levels analyzed by two different methods due to discontinuation of reagents; no disaggregation of data for oral contraceptive vs. DMPAFunded by government |
| Burger et al. [62]Netherlands | Single time point retrospective pharmacokinetic analysisTo characterize factors that influence interpatient variability in EFV concentrations | Sixty-six HIV+ women | Eight hormonal contraceptive userscontaining EFV-containing cART | EFV concentration: No HC: 5.0 mg/l HC: 2.7 mg/l | Strengths: HIV+ womenWeaknesses: Study not designed to look at contraceptive effects; retrospective; single time point; very few hormonal users; population not well described; self-reported hormonal contraceptive use, type not specifiedFunding source not specified |
| Muro et al. [63]Netherlands | Open-label; single-period pharmacokinetic studyTo evaluate factors that influence interpatient variability in single dose NVP half-life | Forty-four healthy nonpregnant women age 18–40 (median 26 years); without hepatitis infection; median weight 64 kg | Seventeen oral contraceptive users 27 nonusersSingle dose of 200 mg of NVP | Seventeen women reported COCs; median time to first undetectable NVP plasma level was 21 days; longer than in 27 women not reporting COC use (14 days; P <0.001)Median half-life of NVP in COC users versus nonusers not significantly different (69.7 vs. 52.8 h; P = 0.053). | Strengths: clearly described population and methods; valid assaysWeaknesses: study not designed to look at contraceptive effects; few hormonal users; healthy women; single dose of single antiretroviral; self-reported hormonal contraceptive useFunding source not specified |
| Frohlich et al. [64]Germany | Open-label; two period pharmacokinetic studyTo investigate the influence of COCs on SQV pharmacokinetic and to assess the potential contribution of CYP3A4 and P-glycoprotein | Eight healthy nonsmoking nonpregnant women with regular menses; mean age 24 years and mean BMI 21; not using any potentially interacting drugs | COC containing GES days 4-25600 mg SQV on days 1 and 22 | No effect of COCs on SQV pharmacokinetics | Strengths: Clearly described population and methods; valid assaysWeaknesses: not randomized; very small sample size; short course of COCs; healthy women; single antiretroviral only given twiceFunded by government |
| Mildvan et al. [53]USA | Open-label, single dose, two period pharmacokinetic studyTo determine the effects of NVP on COC pharmacokinetics and vice versa | Fourteen HIV+ nonpregnant, nonlactating, nonsmoking women; age18–65 (mean age 37); viral load <400; CD4+ cell count > 100 cells/μl; normal renal and hepatic function; no RTV or DLV use | Single dose of COC containing NET on cycle day 1 and 30NVP 200-mg daily on days 2–15; then 200-mg twice daily days 16–29; single dose on day 30cART regimens included IDV; NFV; SQV/RTV | Ten women completed the studyEE AUC ↓29% Cmax unchangedNET AUC ↓18% Cmax unchangedNVP levels similar to historical controls | Strengths: HIV+ clearly described population and methods; valid assaysWeaknesses: small study; only single dose COC; NVP added to current cART regiment; included postmenopausal womenFunded by industry |
| Ouellet et al. [54]Canada | Single dose, single period pharmacokinetic studyTo assess the effects of RTV on EE pharmacokinetics | Twenty-three healthy nonpregnant nonlactating women, 18–45, close to ideal weight; women were postmenopausal, sterilized, practiced abstinence, or had a vasectomized partner | Single dose of COC with 50 μg EE + 1 mg ethynodiol diacetate given on cycle days 1 and 29RTV oral solution from day 15–30, 300 mg q12h on Day 15, 400 mg q12h on Day 16, and 500 mg q12h thereafter | EE Cmax ↓32% AUC ↓41% | Strengths: valid assaysWeaknesses: no progestin levels; nonrandomized; single dose COC; postmenopausal healthy women; nonstandard RTV dosesFunded by industry |
Abbreviations for antiretrovirals and contraceptive steroids defined in Tables 1 and 2.
AUC, area under the curve; Cmax, Peak concentration; Cmin, rough concentration; COC, combined oral contraceptive; DMPA, depot medroxyprogesterone acetate; ECP, emergency contraceptive pill; MPA, medroxyprogesterone acetate; POP, progestin-only pill.