Darunavir

Direction

Reverse or neutral for circulating 25(OH)D (higher after the switch). Not a deficiency finding and no supplementation implication. Baseline low vitamin D in HIV is high (77% below 50 nmol/L in MONET [7]; 86.7% below 30 ng/mL in a French cohort of 2,994 [9]) and is presumed to reflect the underlying condition and prior regimens until excluded (Hard Rule 6).

Component split

Darunavir with ritonavir: all clinical data below. Darunavir with cobicistat: no vitamin D study retrieved. Darunavir without a booster: none (the unboosted form is not used clinically). Ritonavir alone: in vitro mechanism only [11].

Regulatory status

Vitamin D is not mentioned in the US label text read [1]. The SmPC lists osteonecrosis as uncommon with both boosted forms and osteoporosis as uncommon with darunavir/ritonavir, and notes osteonecrosis is multifactorial [2].

Population evidence

MIDAS RCT [6]: adults with HIV RNA under 50 copies/mL on tenofovir/emtricitabine/efavirenz for at least 6 months, randomized 1:1 to continue or switch to darunavir/ritonavir 800/100 mg once daily (ritonavir booster, monotherapy), 48 weeks, 64 analysed; adjusted 25(OH)D difference +3.6 ng/mL favouring darunavir/ritonavir (95% CI 0.6 to 6.6; P = 0.02); renal tubular markers not significantly different. Per-arm means were not in the record read. MONET RCT, vitamin D analysis [7]: 256 adults with HIV RNA under 50 copies/mL randomized to darunavir/ritonavir 800/100 mg once daily alone or with two NRTIs (ritonavir booster); 221 had 25(OH)D measured; 170/221 (77%) below 50 nmol/L at screening; lower baseline levels linked to calendar month (P = 0.0067), black ethnicity (P = 0.013), efavirenz (P = 0.0062) and zidovudine (P = 0.015); rises to week 96 were larger in those who stopped efavirenz or zidovudine. No CIs in the abstract. Pilot RCT, African American adults [10]: 35 randomized, 10 per arm completed; raltegravir plus darunavir/ritonavir arm had a sustained 25(OH)D rise (P = 0.0004), the efavirenz/emtricitabine/tenofovir DF arm did not (P = 0.78); all participants received vitamin D3 and calcium (listing-level read), so the rise is not attributable to the regimen alone. Swiss HIV Cohort population model [8]: 664 participants, 1,397 25(OH)D measurements; a covariate analysis confirmed an effect of darunavir/ritonavir on 25(OH)D, but the abstract gives neither direction nor size. French Dat'AIDS cross-section [9]: 2,994 patients, 2,660 on treatment; efavirenz aOR 1.89 (95% CI 1.45 to 2.47) for deficiency or insufficiency; darunavir was an evaluated drug with no association reported in the abstract (booster not stated).

Class and component evidence (CLASS EXTRAPOLATION)

Cozzolino 2003 [11]: in human hepatocyte (H3B) and monocyte (THP-1) lines, ritonavir, indinavir and nelfinavir suppressed 25-hydroxylase and 1-alpha-hydroxylase in a reversible, dose-dependent way. Darunavir was NOT tested (the study predates its approval), so this is Tier C for darunavir through the ritonavir component and is capped at yellow. Concentrations are not in the abstract, so relevance to booster-dose exposure is unknown.

Mechanism

For darunavir itself: none named. For the ritonavir component: in vitro suppression of vitamin D activation [11]. For efavirenz (the comparator in two trials): induction of vitamin D catabolism is the usual explanation but was not tested in the sources read; the association (aOR 1.89) is documented [9].

Cohort test

Darunavir was in the cohort in [6][7][10] as darunavir/ritonavir (the cobicistat form was in none), so those are Tier A for the regimen but cannot isolate darunavir from ritonavir or from efavirenz withdrawal. [8] and [9] name darunavir without usable effect data. [11] excludes darunavir.

Red/orange hunt

Effect sizes retrieved for MIDAS (+3.6 ng/mL, 0.6 to 6.6) and MONET (counts, P values). Second independent Tier A sources exist (MONET, the pilot RCT) and agree in direction with MIDAS, but the direction is a rise, so red and orange fail on direction, not on missing data. What would change the grade: a cohort of boosted-darunavir starters not switching from efavirenz showing a fall in 25(OH)D or 1,25(OH)2D; none was found.

Counter-evidence held on record

Counter-evidence and confounding Efavirenz is the documented adverse comparator [9], so a rise after switching measures efavirenz withdrawal at least as much as any darunavir effect. Season, ethnicity and zidovudine withdrawal also move 25(OH)D [7]. The pilot RCT supplied vitamin D3 and calcium to everyone [10]. Vitamin D was not measured with cobicistat-boosted darunavir in any study retrieved. Framing caution Do not present darunavir as lowering vitamin D. The human signal is a rise after leaving efavirenz. Nothing here supports vitamin D supplementation because of darunavir; baseline deficiency is common in HIV for other reasons.

Direction

Nutrient (food) to drug, absorption. Hard Rule 4: filed as a timing/administration interaction. It is not a nutrient depletion and implies no change in nutrient intake. The required action is to take the dose with or right after a meal (SmPC: within 30 minutes after finishing a meal), not a separation interval.

Booster

Ritonavir in [37] and the US label; cobicistat in [38][39] and the Prezcobix/Symtuza forms. Both enhancers are in the data.

Regulatory status

US label 2.3 (summarized): must be co-administered with ritonavir, with food [1]. SmPC 4.2: take with cobicistat or low-dose ritonavir within 30 minutes after finishing a meal; "the type of food does not affect the exposure to darunavir" [2]. Didanosine is to be dosed 1 hour before or 2 hours after darunavir/ritonavir (drug-drug timing, label table 10) [1].

Population evidence

Meal types, ritonavir-boosted [37]: open-label randomized crossover, single 400 mg darunavir dose with ritonavir 100 mg twice daily; fasted Cmax and AUClast about 30% lower than after a standard meal; no significant differences among the fed states (standard, high-fat, protein drink, croissant with coffee). Sample size not in the abstract. Darunavir/cobicistat fixed-dose, cobicistat-boosted [38]: phase 1 three-panel crossover (74, 40 and 19 volunteers); high-fat breakfast gave Cmax 2.27-fold and AUC 1.63 to 1.70-fold higher than fasted; fasted/fed CIs not reported in the abstract; bioequivalence of the combined tablet to separate agents within 80 to 125%. Darunavir/cobicistat/emtricitabine/tenofovir alafenamide, cobicistat-boosted [39]: two phase 1 crossover studies (food study 24 volunteers, bioequivalence 96); fasted versus fed darunavir Cmax 45% lower, AUClast 34% lower, AUCinf 30% lower. Secondary compilation [3]: Cmax and AUC about 30% higher fed with ritonavir.

Mechanism

Named in sources only as higher absorption with a meal; the abstracts do not give a mechanism beyond that.

Cohort test

Darunavir was the studied drug in [37][38][39] (healthy volunteers). Exposure was single-dose or short crossover, not chronic therapy.

Red/orange hunt

Not applicable to a deficit: this is a Tier A food effect on absorption with consistent direction across three trials and the label. It cannot exceed yellow under Hard Rule 4.

Counter-evidence held on record

Counter-evidence and confounding Abstract-level figures; CIs and per-arm n for [37] were not retrieved. Healthy-volunteer exposure studies do not show clinical failure from fasting, only lower concentrations. Framing caution The instruction is to take darunavir with a meal as prescribed. It does not mean anyone should add or avoid a nutrient.

Direction

Excess of circulating lipids and glucose (drug or regimen metabolic effect). Outcome entry; no nutrient is implied and no supplementation recommendation follows.

Booster

US label tables are darunavir/ritonavir (800/100 mg once daily in treatment-naive adults; 600/100 mg twice daily in treatment-experienced). SmPC reports separate frequencies for darunavir/ritonavir and darunavir/cobicistat. Ritonavir itself raises lipids, so booster-free attribution is not possible in any source read.

Regulatory status

US label 6 (summarizer): treatment-naive, grade 2 or worse: total cholesterol 23%, triglycerides 3%, LDL 14%, glucose 11%; treatment-experienced: total cholesterol 25% (grade 3 10%), triglycerides 10% (grade 3 7%, grade 4 3%), LDL 14%, glucose 10%, diabetes mellitus 2%; warnings on new-onset or worsened diabetes, hyperglycemia, ketoacidosis and fat redistribution [1]. SmPC 4.8: diabetes mellitus, hypertriglyceridaemia, hypercholesterolaemia and hyperlipidaemia are common with both boosted forms; hyperglycaemia and insulin resistance uncommon with darunavir/ritonavir [2].

Population evidence

Clamp study, ritonavir-boosted [27]: randomized phase 1, healthy volunteers, 28 analysed (arms 10, 9, 9); insulin sensitivity with emtricitabine/tenofovir DF plus darunavir/ritonavir 11.6 +/- 2.5 to 11.3 +/- 2.4 mg/min/kg (not significant); lopinavir/ritonavir arm fell 12.5 to 9.2 (P = 0.037); triglycerides on darunavir/ritonavir 75 (IQR 56 to 95) to 96 (93 to 128) mg/dL (P = 0.009). Darunavir/ritonavir versus atazanavir/ritonavir [28]: phase 4 pilot, treatment-naive, 34 versus 31 enrolled, both with emtricitabine/tenofovir DF; week 12 total cholesterol +20.3 versus +4.6 mg/dL; week 48 no clinically relevant between-arm differences in lipids, fasting glucose or insulin sensitivity. Adipokine substudy, ritonavir-boosted [29]: 234 participants with suppression; leptin +22%, adiponectin +1%, no arm difference between atazanavir/ritonavir, darunavir/ritonavir and raltegravir. Dual therapy [30]: darunavir/ritonavir plus lamivudine versus triple therapy, retrospective, 364 patients; total cholesterol +13 versus +2 mg/dL (P = 0.003), consistent with withdrawal of a lipid-lowering NRTI rather than a darunavir effect (interpretation, not tested). Tenofovir form [32]: darunavir/cobicistat with tenofovir alafenamide versus disoproxil fumarate, 725 randomized; total/HDL cholesterol ratio change 0.20 versus 0.08 (P = 0.036). Cell data [18]: darunavir alone and with ritonavir did not significantly affect insulin signaling or glucose transport in murine and human adipocytes; lopinavir/ritonavir and atazanavir/ritonavir partly impaired it. Large trial context [31]: ACTG A5257, 1,809 participants randomized to ritonavir-boosted atazanavir, darunavir or raltegravir with emtricitabine/tenofovir DF; boosted darunavir and raltegravir had similar tolerability; lipid and glucose numbers were not in the abstract.

Mechanism

Ritonavir-related lipid and insulin effects are the usual explanation; darunavir alone was neutral in adipocytes [18]. Not tested as a nutrient mechanism.

Cohort test

Darunavir (with ritonavir or cobicistat) was in every cohort cited. No source separates darunavir from its booster in people.

Red/orange hunt

Label frequencies and trial effect sizes were retrieved at abstract level. The grade is capped by Hard Rule 2: lipid and glucose changes are outcomes, so no red or orange applies.

Counter-evidence held on record

Counter-evidence and confounding Confounded by booster, tenofovir form, withdrawal of other agents, weight gain on treatment and disease control. Most signals are healthy-volunteer, pilot or abstract-level. Framing caution These are monitoring outcomes of the regimen. Do not describe them as a vitamin or mineral deficiency and do not suggest supplements for them.

Direction

Possible urinary loss of phosphate and potassium through proximal tubular injury (depletion-type disturbance) attributed to tenofovir disoproxil fumarate, with cobicistat raising exposure as a hypothesis. It is a drug-toxicity pathway, not a dietary shortfall (Hard Rule 5), and says nothing about intake.

Booster and attribution

Darunavir with cobicistat in [32][33]; the comparator arms keep darunavir/cobicistat constant and vary only the tenofovir form (TAF versus TDF). Ritonavir-boosted darunavir was in the Fanconi case after the switch [36].

Regulatory status

Potassium, phosphate and magnesium are not mentioned in the US Prezista label text read [1], and hypokalaemia did not appear in the SmPC portions read [2]. The Prezcobix label (which carries the cobicistat creatinine statement) returned 429 and was not read, so no label wording on cobicistat and serum creatinine is cited.

Population evidence

AMBER, week 48 [32]: double-blind randomized, 725 treatment-naive adults, darunavir/cobicistat/emtricitabine/tenofovir alafenamide versus darunavir/cobicistat plus emtricitabine/tenofovir DF; urine protein/creatinine change -22.42 versus -10.34 mg/g (P = 0.033); hip BMD +0.21% versus -2.73% (P less than 0.0001). Phosphate and potassium not in the abstract. Phase 2 [33]: 153 randomized 2:1 (103 TAF, 50 TDF) on a darunavir/cobicistat backbone; creatinine +0.06 (95% CI 0.04 to 0.08) versus +0.09 (0.05 to 0.14) mg/dL; retinol-binding protein/creatinine +9% versus +54% (P = 0.003); beta-2-microglobulin/creatinine -42.0% versus +2.3% (P = 0.002). No phosphate or potassium values in the abstract. Case report [36]: tenofovir DF Fanconi syndrome after 12 years with hypophosphatemia, hypokalemia and vitamin D3 deficiency, resolving over 6 months after a switch to tenofovir alafenamide, emtricitabine, darunavir and ritonavir (abstract level; attributed to tenofovir DF). Case report [21]: severe lactic acidosis with tenofovir DF and cobicistat in a darunavir/cobicistat regimen; attributed to tenofovir with cobicistat as a possible contributor.

Mechanism

Proximal tubular injury from tenofovir DF; cobicistat inhibits renal transporters and raises tenofovir exposure (hypothesis in [21]; not tested in the abstracts read).

Cohort test

Darunavir/cobicistat was in the cohorts in [32][33]; the contrast is TAF versus TDF, so darunavir itself is a constant. EMERALD and week-96 extension records were seen at listing level only [34].

Red/orange hunt

No Tier A source measured phosphate or potassium directly in darunavir users in what was retrieved, and the available tubular differences are attributable to the tenofovir form. Orange not available.

Counter-evidence held on record

Counter-evidence and confounding Attribution to darunavir is contradicted by the design: an unchanged darunavir/cobicistat backbone with better markers on TAF. Framing caution This is a medication safety signal for tenofovir disoproxil fumarate regimens. It does not mean phosphate or potassium supplementation is warranted because of darunavir.

Direction

Bone outcome (BMD and turnover markers): improvement after leaving tenofovir DF, loss on tenofovir DF. Not a calcium or vitamin D finding.

Booster

Ritonavir in [6][35][10]; cobicistat in [32][33].

Regulatory status

Osteonecrosis is listed in the US label (less common reactions, treatment-naive) [1] and as uncommon in the SmPC with both boosted forms, with osteoporosis uncommon with darunavir/ritonavir; the SmPC calls osteonecrosis multifactorial [2]. Calcium is not mentioned in the label text read.

Population evidence

MIDAS [6]: 64 analysed; femoral neck BMD +2.9% versus -0.003% and lumbar spine +2.6% versus +0.008% on darunavir/ritonavir versus continuing tenofovir/emtricitabine/efavirenz (P less than 0.05); bone turnover markers improved. RADAR [35]: randomized, 85 treatment-naive, darunavir/ritonavir in both arms with raltegravir or tenofovir DF/emtricitabine; subtotal BMD change +9.2 versus -7 (P = 0.002, units as stated in the abstract), CTX +0.04 versus +0.24 ng/mL (P = 0.001), P1NP +3.59 versus +30.09 ng/mL (P = 0.023); virologic response 62.5% versus 83.7% (P = 0.045). AMBER [32] and phase 2 [33]: BMD better on tenofovir alafenamide than disoproxil fumarate with darunavir/cobicistat constant (hip -0.84% versus -3.82%, P less than 0.001; spine -1.57% versus -3.62%, P = 0.003 in [33]).

Mechanism

Tenofovir DF bone effect; efavirenz and vitamin D catabolism for the MIDAS comparison. No darunavir-specific bone mechanism is named in the sources read.

Cohort test

Darunavir was in every arm of [35][32][33], so the contrasts do not isolate darunavir; in [6] the comparator contains tenofovir DF and efavirenz.

Red/orange hunt

Effect sizes retrieved. They do not point to a darunavir-attributable bone loss and calcium/PTH were not measured, so no deficit grade is available.

Counter-evidence held on record

Counter-evidence and confounding Switch trials favour the new regimen because tenofovir DF and efavirenz were removed. Framing caution No calcium or vitamin D supplementation is implied.

Direction

Herbal product to drug (loss of antiviral exposure). Hard Rule 4: not a depletion. A separation interval cannot fix an inducer; the labels contraindicate co-use.

Booster

Boosted forms with ritonavir or cobicistat (SmPC) and darunavir/ritonavir (US label).

Regulatory status

US label section 4 lists St John's wort as contraindicated [1]. SmPC 4.3 and 4.5: boosted darunavir must not be used with products containing St John's wort; if already taking it, stop and check viral load if possible [2].

Population evidence

Label-based (Tier A regulatory); no trial retrieved.

Mechanism

Induction of CYP3A lowers darunavir and enhancer levels (label text).

Counter-evidence held on record

Framing caution Do not combine with St John's wort; discuss any herbal product with a pharmacist.

Limits on what this shows

• Why no grade: Dual condition met: no Tier B or C evidence and no circulating claim. The frequent low zinc and selenium in treated HIV [53] is a baseline-prevalence and indication context that belongs in monitoring thinking, not a darunavir attribution (Hard Rule 6). Governance: none assigned.

What was searched (first pass)

PubMed listing "darunavir ferritin OR iron OR zinc OR copper OR selenium" [42] (11 results, 10 shown): none measured zinc status. Zinc hits: an open-label COVID-19 trial of ivermectin plus zinc sulfate versus hydroxychloroquine plus darunavir/ritonavir (113 participants; zinc was a treatment, not measured), and a protein crystallography paper in which zinc acetate was a buffer component [42]. Label and SmPC text read: no zinc statement [1][2]. Consumer pages [44][45]: no nutrient claim.

What was searched (repair pass, 2026-10-10)

PubMed listing "darunavir zinc OR darunavir selenium serum" [48] (8 records): no record measured zinc, selenium or copper in darunavir users; zinc was a treatment in a COVID-19 trial, a database name in an in-silico screen, and a reagent in three assay or crystallography papers. Web search for serum zinc, copper or selenium in HIV patients on antiretroviral therapy with darunavir. Opened and cohort-tested: a four-arm zinc and selenium supplementation pilot in 60 Mexican men on long-term ART [52] (regimens efavirenz, emtricitabine, tenofovir DF, abacavir/lamivudine and "others"; darunavir and protease inhibitors not named; baseline serum zinc about 82 to 90 and deficiency 19%); a zinc and selenium intake and serum study in 124 men with HIV (serum subcohort 45; 23.9% zinc deficient, 65.9% selenium deficient) [53] (drugs not named, darunavir and protease inhibitors not mentioned); and the PEARLS pre-ART selenium case-cohort [54] (atazanavir arm, 83 participants, median selenium 76 ug/L, 58% deficient, versus 84 and 85 ug/L with efavirenz arms, P = 0.61 across arms; atazanavir, not darunavir, and baseline before ART). None qualifies as a darunavir cohort, so none supports a yellow class extrapolation for darunavir.

Channel outcomes

Genuine gap on the channels that opened; same tooling limits as copper (PubMed Central blocked, copper phrasing rate-limited).

Limits on what this shows

• Why no grade: Dual condition met on the sources actually read: no Tier B or C evidence of a darunavir-linked change and no circulating claim. High baseline hyperhomocysteinemia and low B12 in treated HIV [50] is indication-linked context (Hard Rule 6), not a drug attribution. Governance: none assigned.

What was searched (first pass)

PubMed listing "darunavir vitamin deficiency" [43] (5 records: four vitamin D studies and a Fanconi case report with vitamin D3 deficiency): none reported B12, folate, vitamin A, E, K, carnitine or selenium. Label and SmPC text read: no B12 or folate statement [1][2]. No consumer claim found [44][45].

What was searched (repair pass, 2026-10-10)

PubMed listing "darunavir vitamin B12" [46] (2 records, completed this time): a COVID-19 therapeutics review in which vitamin B12 is a natural compound and darunavir a drug, and a three-case olfactory dysfunction report in COVID-19; neither reports B12, folate or homocysteine in darunavir users. "darunavir folate" and "darunavir homocysteine" errored again (client error). Web search (two phrasings: B12, folate or homocysteine with HIV regimens naming darunavir, atazanavir, lopinavir; darunavir with B12, homocysteine or folate in plasma). Opened and cohort-tested: Uccelli 2006, 98 patients on stable HAART exposed to protease inhibitors [49] (cross-sectional; normal folate protective against hyperhomocysteinemia, multivariable OR 0.24, 95% CI 0.06 to 0.94; length of protease inhibitor exposure not associated; darunavir not named, and the study predates darunavir approval, so it is Tier C for darunavir and also shows no protease inhibitor effect on folate or homocysteine); and a 57-patient Chieti-Pescara cross-sectional study of homocysteine, B12 and vitamin D with neurocognition [50] (homocysteine raised in 70.2%, B12 low in 14.0%; the authors report no significant relation between therapy and vitamin levels, with no figures; darunavir not named, and no folate result).

Channel outcomes

Genuine gap on the sources opened; residual tooling limit: the folate and homocysteine PubMed phrasings failed twice across passes, so a darunavir-specific record cannot be fully excluded. The class study found no protease inhibitor effect, so it gives no basis for a class extrapolation yellow.