Voriconazole

Limits on what this shows

• Null findings: In the Thompson cohort no voriconazole patient had radiographic periostitis or exostoses (0 of 9; 5 imaged), so a raised fluoride does not always come with bone disease [13]. In that cohort fluconazole and posaconazole were not associated with hyperfluorosis [13], and in the Zurich cohort neither posaconazole nor itraconazole raised fluoride [12]. In the index case of the 2021 review, plasma fluoride was 0.4 mg/L (stated normal 0.2 to 3.2) at presentation, and serum fluoride became undetectable after stopping, so symptoms can occur with a value inside the quoted range [15].
• At-risk subgroups: Courses of several months or more (onset median 6 months [14]), transplant recipients, hematology patients, renal insufficiency [12]. Also children: two paediatric patients are in the review [15] and a teenager with leukemia is a case report (listing level) [24].

Direction

Excess (retention) of fluoride from the drug itself. Periostitis, fluorosis and bone pain are clinical outcomes of that excess and are not nutrient deficits (Hard Rule 2). Fluoride was measured directly in the cohorts below. Supplemental fluoride would be the hazard, not the remedy (Hard Rule 5).

Regulatory status

US label, warning 5.12: "Fluorosis and periostitis have been reported during long-term VFEND therapy"; the label says to discontinue if skeletal pain and radiologic findings compatible with fluorosis or periostitis develop [1]. EU SmPC section 4.4: "Non-infectious periostitis with elevated fluoride and alkaline phosphatase levels has been reported in transplant patients" [2]. US label adverse reaction table: alkaline phosphatase increased in 59 of 1,655 (3.6%) therapeutic-study patients [1].

Population evidence

Wermers 2011 [11]: 10 adult post-transplant patients on voriconazole for at least 6 months against 10 post-transplant controls not on it; plasma fluoride 14.32 +/- 6.41 umol/L against 2.54 +/- 0.67 umol/L (P <.001); half of the voriconazole group had periostitis (exostoses in 2); alkaline phosphatase fell after voriconazole was stopped in the patients with periostitis; no reference range in the abstract. Gerber 2012 [12]: retrospective, 43 hematologic patients (voriconazole 20, posaconazole 8, itraconazole 4, controls 11); serum fluoride was raised only in the voriconazole group (median 156.5 ug/L, IQR 96.8; normal under 30 ug/L; P <.001 against every comparison group), about 5-fold above normal; bone pain and radiologic periostitis occurred only on long-term voriconazole and improved in all after stopping; clinically relevant skeletal disease was linked to renal insufficiency and to fluoride about 10-fold above normal. Thompson 2012 [13]: 29 patients with coccidioidomycosis on long-term triazoles (voriconazole 9, fluconazole 9, posaconazole 5, itraconazole 6); mean fluoride 9.17 umol/L (range 1.83 to 21.0) on voriconazole against 2.98 (0.97 to 4.53) on fluconazole and 4.06 (3.00 to 5.34) on posaconazole, normal range quoted as 1 to 4 umol/L, two voriconazole patients above 20 umol/L. Case synthesis [14]: 44 patients from 34 reports, median onset 6 months after starting, diffuse skeletal pain in 68.28%, all resolved after stopping and 18 (52.94%) improved within one week; a review of 89 reported cases lists solid organ transplant 45%, immunocompetent hosts 21.3% and hematologic malignancy 20.3% [15]. Moon 2014 [31] (a fourth cohort, opened in the repair pass; all patients were on voriconazole, so the comparison is periostitis against no periostitis rather than against unexposed people and it is not counted among the three above): in a retrospective study of 195 voriconazole-treated patients from a contaminated-methylprednisolone outbreak, 28 had both bone scan and plasma fluoride; mean plasma fluoride was 12.78 +/- 0.96 umol/L (n = 21, scan-positive periostitis) against 3.61 +/- 1.29 umol/L (n = 7, scan-negative) (P <.001; laboratory reference below 5.26), alkaline phosphatase 273 +/- 35.6 against 117 +/- 15.7 IU/L (P =.020), daily dose 780 +/- 43 against 450 +/- 82 mg (P <.001); a cutoff above 8 umol/L gave sensitivity 0.95 (95% CI 0.76 to 0.99) and specificity 1.0 (95% CI 0.59 to 1.0); 29 of 195 (15%) were diagnosed with periostitis and pain improved in 89% after dose reduction or stopping (values +/- are standard errors as printed).

Mechanism

Named: the drug is a fluorinated triazole (Thompson calls it trifluorinated [13]); reviewers attribute the periostitis to its fluoride content, with fluoride depositing in the bone matrix [15]. Impaired fluoride clearance in renal insufficiency modified risk in the Zurich cohort [12]. A case report links the syndrome to a CYP2C19 ultrarapid metabolizer genotype and describes urinary alkalinization to raise fluoride excretion (one patient, hypothesis only) [16].

Counter-evidence held on record

Counter-evidence and confounding All three cohorts are small, single-center and observational, with different assays and units (umol/L and ug/L), so the effect size is stated per study and not pooled. The transplant and hematology settings carry renal impairment and many co-medications that also affect bone and alkaline phosphatase. The Thompson letter found raised fluoride without bone disease in 9 patients. The EU label limits the statement to transplant patients. Periostitis itself can mimic other bone disease and is an outcome, not a nutrient entry. Framing caution Excess-direction signal. Not a deficiency, and no fluoride-containing product is indicated. Bone pain on long-term voriconazole is a reason to contact the prescriber, who may check fluoride and alkaline phosphatase or review the drug.

Limits on what this shows

• Null findings: Study 305 had no hypokalemia events in either the voriconazole (0 of 200) or the fluconazole (0 of 191) arm [1]. In the registration trials the comparator drug amphotericin B produced about twice the potassium-lowering rate [1]. An FAERS analysis of voriconazole in hematologic malignancy (605 reports) did not list hypokalaemia among its positive signals [10]. CLASS EXTRAPOLATION context: a triazole FAERS analysis ranked hypokalemia among strong class signals, with the voriconazole-specific ratio not in the abstract, so it is not used to grade [9]. The Lancet 2021 randomized trial of posaconazole against voriconazole (575 treated; 287 on voriconazole) was opened at abstract level in the repair pass [30]: treatment-related adverse events occurred in 40% on voriconazole against 30% on posaconazole, and the treatment-related events above 3% were raised liver enzymes, hallucination, nausea and blurred vision, with hypokalaemia not among them. Per-event counts and all-cause hypokalaemia rates are in the full text and appendix, which were not opened, so this is a soft null on drug-attributed potassium events and not an incidence. A 2021 single-center study (165 neutropenic patients screened) reported lower serum potassium in an FMO3 E308G genotype group than in wild type (p = 0.028), absolute values not in the listing summary [25].
• At-risk subgroups: Women, intravenous dosing, concomitant antibiotics and higher N-oxide to parent ratio [5]; children (11% in label) [1]; people with other QT-prolonging drugs or hypokalemia at baseline [1][7].

Direction

Disturbance (serum potassium falls), not a dietary deficiency, and not a reason to raise potassium intake (Hard Rule 5). The separate direction, low potassium amplifying voriconazole QT toxicity, is a nutrient-state modifier of drug toxicity (Hard Rules 2 and 4); QTc prolongation and torsades are drug outcomes, not deficiencies.

Regulatory status

US label, warnings 5.2 and 5.10: "Rigorous attempts to correct potassium, magnesium and calcium should be made before starting and during voriconazole therapy" and electrolyte disturbances such as hypokalemia, hypomagnesemia and hypocalcemia "should be corrected" before and during therapy [1]. Table 4: hypokalemia in 26 of 1,655 (1.6%) therapeutic-study patients; in studies 307/602 and 608, 3 of 468 (0.6%) on voriconazole against 36 of 185 (19.5%) on amphotericin B and 16 of 131 (12.2%) on amphotericin B then fluconazole; none of 200 on voriconazole and none of 191 on fluconazole in study 305. Laboratory tables: potassium below 0.9 times the lower limit of normal in 30 of 181 (16.6%) on voriconazole against 70 of 178 (39.3%) on amphotericin B (invasive aspergillosis), and 43 of 258 (16.7%) against 35 of 118 (29.7%) (candidemia). Pediatric: hypokalemia in 11 of 105 (11%) [1]. The EU SmPC lists hypokalaemia among metabolic adverse reactions and advises monitoring and correcting electrolytes [2].

Population evidence

Cheng 2024 [5]: prospective, 272 adults with voriconazole therapeutic drug monitoring (414 trough concentrations); hypokalemia in 18.0% (48 of 266); 81.2% (39 of 48) developed it within 14 days and 56.2% (27 of 48) within one week; independent risk factors were sex (more common in females), combined antibiotic use and the N-oxide to parent concentration ratio; it was more frequent with intravenous dosing; the hypokalemia threshold and odds ratios are not in the abstract (read via PubMed record and OpenAlex record, which agree). Li 2026 [6] (full text): self-controlled prescription sequence symmetry analysis in 364 children with hematologic disease; the prespecified "weak positive" control, potassium below 3.0 mmol/L, gave an adjusted sequence ratio of 2.86 (95% CI 1.32 to 6.18). Mycoses 2024 [25]: propensity-matched cohort in cryptococcal meningitis, 47 patients on voriconazole plus flucytosine against 92 on amphotericin B deoxycholate plus flucytosine; both regimens significantly lowered serum potassium over the first 2 weeks (all p of.010 or less; no means in the abstract; OpenAlex record); flucytosine is a confounder. Gueta 2017 [25]: retrospective, 54 hemato-oncology patients paired on and off voriconazole; mean QTc 448.0 against 421.8 ms (p = 0.002), and low serum potassium was independently associated with QTc prolongation (listing summary). Barreto 2019 [7]: 316 adults on voriconazole prophylaxis for leukemia or high-risk MDS; hypokalemia was associated with QTc prolongation (HR 3.15, P =.003; 37 patients, 16.2%, prolonged), a risk-modifier finding, not a potassium-loss estimate.

Mechanism

No mechanism for voriconazole potassium loss is established in anything read. The pseudohyperaldosteronism route that explains posaconazole and itraconazole hypokalemia is not signalled for voriconazole: in an FAERS and randomized-trial meta-analysis posaconazole carried a hypokalemia risk ratio of 10.41 (95% CI 2.00 to 54.27, I2 = 0%) against voriconazole as comparator [8], and cell-assay work found no comparable steroid-pathway effect for the other azoles (listing level) [25]. The Cheng finding that the N-oxide to parent ratio predicts hypokalemia suggests an exposure or metabolism link but is not a mechanism.

Counter-evidence held on record

Counter-evidence and confounding The strongest controlled data are active-comparator: amphotericin B is a known potassium-wasting drug, so the voriconazole arm looking better than it cannot show that voriconazole has no effect, and cannot attribute the remaining events to voriconazole either. The Cheng cohort has no unexposed arm and enrolled hospitalized patients with chemotherapy, diarrhea, diuretics and antibiotics. The Li 2026 control analysis is a design check inside a paediatric heme cohort. A paediatric trial and adult cohorts at listing level were not opened. Because no named mechanism exists, red was not met; orange is bounded by setting and timing. Framing caution Monitoring prompt for people with arrhythmia risk, not a potassium deficiency claim and not a reason to take potassium. Potassium changes (including replacement, which can itself be unsafe in kidney disease) belong with a prescriber.

Limits on what this shows

• Null findings: The dose-response check was flat (no trend across weight-based dose tertiles: 16.2%, 10.7%, 10.2%; P = 0.437) [6]. The authors cite two FAERS analyses that did not detect a voriconazole-hypomagnesemia signal [6]; the FAERS study of voriconazole in hematologic malignancy opened here listed no hypomagnesaemia signal [10]. No adult magnesium cohort for voriconazole was found: the PubMed listing "voriconazole magnesium" returned 13 records, of which one (Li 2026) measured magnesium [26]. In the Lancet 2021 trial abstract [30] hypomagnesaemia was not among the treatment-related events above 3% in the voriconazole arm (soft null; per-event counts not opened). Moon 2014 [31] (195 voriconazole-treated patients in a methylprednisolone outbreak) reported creatinine, ALT, bilirubin and trough values and no magnesium or potassium measure.
• At-risk subgroups: Children with hematologic malignancy or related disease treated for weeks (onset median 28 days); the loop-diuretic stratum had the highest ratio (8.30) but small cells; co-prescribed PPIs (45.6%), loop diuretics (34.1%) and aminoglycosides (51.1%) were common [6].

Direction

Disturbance (serum magnesium falls), not a dietary deficiency and not a reason to supplement without a prescriber (Hard Rule 5). A separate direction, low magnesium raising voriconazole QT risk, is a nutrient-state modifier of drug toxicity (Hard Rule 4) and is reported in the Barreto cohort below.

Regulatory status

US label lists hypomagnesemia in the electrolyte warning (5.10) with potassium and calcium, in the list of less common adult reactions (under 2%, no percentage) and as a pediatric adverse reaction in 5 of 105 (5%) [1]. The EU SmPC advises monitoring and correcting hypomagnesaemia but does not list it in the section 4.8 table as read [2].

Population evidence

Li 2026 [6] (full text opened on Frontiers; PubMed and OpenAlex records agree): single-center Children's Hospital of Chongqing, inpatients November 2013 to December 2024; 536 children on voriconazole, 364 (67.9%) with serum magnesium measured; 58 patients in the post-exposure sequence against 8 in the reverse sequence; crude sequence ratio 7.25, adjusted sequence ratio 5.76 (95% CI 2.75 to 12.06); 66 of 532 (12.4%) fell below 0.60 mmol/L within 90 days; median onset 28 days (IQR 20 to 43), 55.2% within 30 days; median serum magnesium 0.79 mmol/L; threshold sensitivity 4.67 (below 0.70), 5.92 (0.65), 5.76 (0.60), 5.97 (0.55). After correcting for greater testing after voriconazole start (1.73-fold more magnesium measurements) the ratio was 3.33 (95% CI 1.59 to 6.98). Prespecified controls behaved as expected: sodium below 130 mmol/L 1.18 (0.58 to 2.42, negative), bilirubin above 34 umol/L 1.05 (0.47 to 2.36, negative), potassium below 3.0 mmol/L 2.86 (1.32 to 6.18, positive), ALT above 80 U/L 1.72 (1.21 to 2.44, positive). Strata: with PPI 5.00 and without 5.78; with aminoglycoside 4.58 and without 5.80; with loop diuretic 8.30 and without 2.53 (0.95 to 6.74, includes 1). Barreto 2019 [7]: in 316 adults on voriconazole prophylaxis, hypomagnesemia was associated with QTc prolongation (HR 6.47, P =.007; no CI or n in the abstract); this is the nutrient-to-toxicity direction and not a magnesium-loss estimate.

Mechanism

None established for voriconazole. The authors state that "the molecular mechanism underlying the observed voriconazole-hypomagnesemia association remains unknown" and discuss PPI, calcineurin inhibitor and aminoglycoside pathways as other causes [6]. Orange does not require a named mechanism; red does.

Counter-evidence held on record

Counter-evidence and confounding Prescription sequence symmetry analysis controls for fixed patient traits but not for time-varying factors such as chemotherapy intensity, disease severity, nutrition or concurrent nephrotoxic drugs, and the authors say it may overestimate the ratio. Magnesium was measured more often after voriconazole started, and the corrected ratio is exploratory. Amphotericin B was used in 51 children and platinum agents in 4, too few to stratify, yet both waste magnesium. There was no voriconazole drug-level data. One center, one country, and the reverse sequence may partly reflect admission testing. No second independent Tier A cohort with magnesium measured was found, so red was not met. Framing caution Monitoring prompt, not a magnesium deficiency claim. The authors suggest weekly serum magnesium during the first 30 days; any replacement decision belongs with a prescriber, particularly in kidney disease.

Limits on what this shows

• Null findings: The Li 2026 negative control was null (above). Hyponatremia is in the label only as a rare reaction [1].
• At-risk subgroups: Adults over about 47 years, long courses and high troughs in one cohort [5].

Direction

Disturbance (serum sodium falls), a dilutional or hormonal effect and not a dietary deficiency; it never implies that salt intake should increase (Hard Rule 5).

Regulatory status

US label: hyponatremia is in the less common (under 2%) adult reaction list without a percentage [1]. The EU SmPC lists hyponatraemia under metabolism and nutrition disorders as read through a summarizing tool; the frequency band was not legible [2].

Population evidence

Cheng 2024 [5]: hyponatremia in 21 of 266 adults (7.9%); 71.4% (15 of 21) were over 47 years old; the hyponatremia group had longer voriconazole use and 47.6% (10 of 21) had supratherapeutic troughs (above 5.0 ug/mL). Li 2026 [6]: sodium below 130 mmol/L was the prespecified negative control, adjusted sequence ratio 1.18 (95% CI 0.58 to 2.42). FAERS [10]: adults with hematologic malignancy, inappropriate antidiuretic hormone secretion n = 4, ROR 9.7 (3.61 to 26.04).

Mechanism

Not established. SIADH (inappropriate antidiuretic hormone secretion) appears as a pharmacovigilance signal [10] but no source read tests it for voriconazole.

Counter-evidence held on record

Counter-evidence and confounding No unexposed or active-comparator group; hospitalized patients with many other causes of low sodium. Orange needs a Tier A source showing a real, bounded effect; the one positive cohort has no comparator and the one controlled analysis was null. Framing caution Disturbance signal. Not a reason to change salt or fluid intake; low sodium symptoms warrant prescriber assessment.

Limits on what this shows

• Null findings: None found for calcium or phosphate; no study compared them with a control.
• At-risk subgroups: Children treated for hematologic disease (label and FAERS) [1][10].

Direction

Mixed and unresolved: disturbances in both directions. Not dietary deficiencies; no supplement implication (Hard Rule 5).

Regulatory status

US label pediatric table (N = 105): hypocalcemia 6 (6%), hypophosphatemia 6 (6%), hypomagnesemia 5 (5%), hypoalbuminemia 5 (5%); adults: hypocalcemia, hypophosphatemia and hypercalcemia are in the under-2% list without percentages [1].

Population evidence

FAERS, hematologic malignancy [10]: children, hypercalcaemia n = 5, ROR 29.11 (95% CI 11.1 to 76.3); musculoskeletal pain n = 6, ROR 36.45 (14.84 to 89.51); adults, blood alkaline phosphatase increased n = 9, ROR 7.08 (3.66 to 13.66). These are reporting signals, not measured calcium or phosphate changes in a cohort.

Mechanism

Not established for calcium or phosphate. The hypercalcemia and musculoskeletal pain signals sit beside the fluoride and periostitis entry but no source links them; raised alkaline phosphatase belongs to that fluoride entry.

Counter-evidence held on record

Counter-evidence and confounding Paediatric label percentages are all-cause adverse events in hospitalized children, not attributed. FAERS counts are small (n = 5 and 6) and reporting-biased. Framing caution Context only. Not a calcium or phosphate deficiency claim; no supplement is implied.

Limits on what this shows

• Null findings: Cell-assay work on adrenal steroid biosynthesis found posaconazole and itraconazole, and not the other azoles, to be implicated in pseudoaldosteronism (listing level) [25].
• At-risk subgroups: People receiving systemic, inhaled or intranasal corticosteroids [1][17].

Direction

Hormone suppression and raised corticosteroid exposure through CYP3A4 inhibition. Not a nutrient deficiency (Hard Rule 2); included so the reader sees it is not one.

Regulatory status

US label 5.8: "Reversible cases of azole-induced adrenal insufficiency have been reported in patients receiving azoles, including VFEND"; Cushing syndrome with and without subsequent adrenal insufficiency in patients on voriconazole with corticosteroids; monitor for adrenal dysfunction [1]. EU SmPC: reversible adrenal insufficiency, and in patients on azoles without corticosteroids it is related to direct inhibition of steroidogenesis [2] (quoted fragment).

Population evidence

Duman 2017 [17]: case report and systematic review; probable drug-induced Cushing syndrome followed by adrenal insufficiency on oral voriconazole with intranasal mometasone and inhaled fluticasone; abstract read through an OpenAlex record, no numbers. FAERS [10]: "endocrine disorders" was a system-organ signal in adults with hematologic malignancy. The PubMed listing "voriconazole adrenal insufficiency" returned 6 records, all read at snippet level [26].

Mechanism

Named: CYP3A4 inhibition raising exogenous steroid levels, and direct inhibition of steroidogenic enzymes [1][2][17].

Counter-evidence held on record

Counter-evidence and confounding Case-level evidence for the combination; no controlled cortisol measurements with voriconazole alone were opened. Not a nutrient, so it cannot reach a nutrient tier. Framing caution Hormone effect, not a nutrient deficiency. Signs such as persistent fatigue, low blood pressure or vomiting warrant clinical assessment.

Limits on what this shows

• Null findings: The label text read names no calcium, magnesium, iron or antacid interaction [1].
• At-risk subgroups: Anyone taking oral tablets or suspension; tube-fed patients were not covered in the text read.

Direction

Food to drug (absorption). Hard Rule 4: filed as a timing interaction with a separation interval, not as a depletion.

Regulatory status

US label section 2.1: "Administer VFEND Tablets or Oral Suspension at least one hour before or after a meal" [1]. EU SmPC 5.2: "When multiple doses of voriconazole are administered with high fat meals, Cmax and AUCtau are reduced by 34%" (quoted fragment through a summarizing tool); section 4.2: tablets "at least one hour before, or one hour following, a meal" [2]. US label contraindicates use with drugs and herbal products that induce CYP2C19, CYP2C9 and/or CYP3A4 [1].

Population evidence

Label-level pharmacology only; no supplement-specific, mineral, antacid or dairy interaction study of voriconazole was retrieved.

Mechanism

Reduced absorption with food; oral bioavailability is 96% [2].

Counter-evidence held on record

Counter-evidence and confounding The US label pharmacokinetic section 12.3 was not in the text read; food magnitude comes from the EU SmPC through a summarizing tool. Framing caution Timing interaction only: one hour before or one hour after a meal. No nutrient is removed by the drug.

Limits on what this shows

• Null findings: Rates on voriconazole were below the active comparator [1].
• At-risk subgroups: Patients with chemotherapy-related nausea or diarrhea.

Direction

Changed intake and loss (mechanism class H). An indirect, non-specific route; a competing explanation for electrolyte entries.

Regulatory status

US label Table 4: nausea 89 of 1,655 (5.4%), vomiting 72 of 1,655 (4.4%) in therapeutic studies; in studies 307/602 and 608, nausea 3.8% and vomiting 3.2% on voriconazole against 15.7% and 9.7% on amphotericin B; diarrhea is in the under-2% list [1].

Population evidence

No intake, weight or nutritional study of voriconazole was retrieved.

Mechanism

Gastrointestinal adverse effects; not specified.

Counter-evidence held on record

Counter-evidence and confounding All-cause adverse events in very ill patients; most nausea and vomiting in this population arise from chemotherapy and disease. Framing caution Context flag. Persistent vomiting or poor intake is a reason for clinical review, not a nutrient claim.

Limits on what this shows

• Why no grade: Dual condition met on what was reachable: no Tier B or C source was seen that supports a vitamin D relationship for voriconazole (the single measured case is normal; the gene-variant study runs from vitamin D pathway genotype to drug concentration, the wrong direction for a depletion claim, and measured no vitamin D status; the hypercalcemia case levels predate the drug), and no circulating vitamin D claim was found to rebut. Bone pain and periostitis on long-term voriconazole are fluoride outcomes and are not vitamin D deficiency or low bone density (Hard Rule 2). A later revision should re-run the vitamin D and bone density listings.

What was searched

PubMed "voriconazole vitamin D bone" returned 0 results; "voriconazole bone density" returned 4 results, all read at listing-summary level [29]: a periostitis case synthesis (Li 2022), a pharmacogenomics review, a tamoxifen pharmacokinetic model and a fungal infection case, none with a bone density, 25-hydroxyvitamin D, PTH or fracture measure in voriconazole users. "Voriconazole vitamin D" (plain and in format=pubmed), "voriconazole fracture", "voriconazole osteoporosis" and "voriconazole 25-hydroxyvitamin D" were refused (HTTP 429 or a verification page) and not read. The periostitis listing (26 results, 16 read) [24] and the reviews and cohorts opened [11][12][13][14][15] concern fluoride excess and periostitis, and none reports 25-hydroxyvitamin D in a cohort. In the 2021 review the index patient had "Serum ionized calcium, vitamin D levels, and parathyroid hormone tests were all within normal limits" [15] (a single case, summarizing tool). Label text read: no vitamin D statement [1][2]. FAERS signals for hypercalcaemia and musculoskeletal pain in children [10] are outcomes, not vitamin D measures. Repair pass, all read at listing-summary level [35]: PubMed "voriconazole vitamin D" (4 results: a vitamin D pathway gene-variant study of voriconazole, itraconazole and posaconazole plasma exposure in 636 patients, which tests genotype as a predictor of drug concentration and reports no vitamin D status; an analytical workflow paper; a cystic fibrosis review; and a histoplasmosis hypercalcemia case with 25-hydroxyvitamin D 14.4 ng/mL and PTH below 6.3 pg/mL measured before voriconazole was started), "voriconazole osteoporosis" (1 result, a 2004 new-drug overview with no patient data), "voriconazole CYP24A1" (1 result, the same gene-variant study) and "voriconazole fracture" (33 results, 20 read: fractures appear as injuries or sites of fungal infection in case reports, none with 25-hydroxyvitamin D, PTH or bone density). The 195-patient Moon 2014 cohort [31] reported fluoride, alkaline phosphatase, creatinine and liver tests and no vitamin D, calcium, PTH or bone density.

Channel outcomes

The plain vitamin D, fracture and osteoporosis listings were refused; the bone density listing returned only 4 unrelated-to-measurement records. The repair pass then read the plain vitamin D, osteoporosis, CYP24A1 and fracture listings (the 13 unread fracture records were not read). Open-web search remained unavailable. Result: the vitamin D, osteoporosis and bone density PubMed listings are genuine absences of any cohort measuring 25-hydroxyvitamin D, PTH or bone density in voriconazole users; the fracture listing is partly read, and web-search circulating-claim checks remain tooling-limited. A later revision could read the remaining fracture records and the 25-hydroxyvitamin D phrasing.