Reference standards were obtained from Cayman Chemical (Ann Arbor, USA), the National Measurement Institute (NMI) (Lindfield, Australia), and Toronto Research Chemicals (North York, Canada). Solvents and reagents, including formic acid, ammonium formate, and methanol were of analytical or HPLC grade (Tedia, Fairfield, USA). Ultrapure water (18 MΩ·cm) was produced using a Milli-Q purification system (Millipore, Burlington, MA, USA). Mitra® devices equipped with volumetric absorptive microsampling (VAMS®) tips (30 µL) were purchased from Alllcrom (São Paulo, Brazil). The medications Neosaldina® (300 mg dipyrone, 30 mg isometheptene mucate, and 30 mg anhydrous caffeine), Azopt (brinzolamide 1%), and Cosopt (dorzolamide hydrochloride 2% and timolol maleate 0.5%) were obtained from local retailers.

The study protocol was approved by the Ethics Committee of the Federal University of Rio de Janeiro (protocol 40915120.0.0000.5257). Written informed consent was obtained from all participants. Healthy adults of both sexes, aged 20–43 years, were recruited.

Neosaldina® was administered orally to volunteers (n = 10) as a single dose of 30 mg isometheptene mucate. For diuretics, Azopt® (brinzolamide) and Cosopt® (dorzolamide) were administered to volunteers as either ocular (n = 2 each) one drops in each eye, or two drops orally (n = 2 each). Oral administration was conducted using 2 mL of yogurt as a vehicle.

Capillary blood and urine samples were collected at baseline (0 h) and at 2, 5, 7, 24, and 48 h post dose. These time points were selected to capture the absorption phase, peak systemic exposure, and early elimination of the investigated substances. For diuretics, additional late collections were performed at 7 days and 38 days after administration to evaluate long-term urinary detectability, considering their known prolonged systemic retention. Samples from chronic patients undergoing ophthalmic treatment with Azopt® or Cosopt® (one drop twice daily, as prescribed) were collected to provide comparative data representative of repeated therapeutic exposure. Venous blood was drawn 2 h post-dose into K2EDTA-coated tubes (BD Vacutainer®) to determine hematocrit (HCT) and to obtain plasma (for isometheptene) and whole blood (for diuretic analysis), enabling comparison between venous and capillary blood concentrations. Capillary blood samples were collected from the fingertip using Mitra® VAMS devices following the manufacturer's instructions. After samples collection, urine specific gravity was measured and used for concentration adjustment when applicable, in accordance with routine anti-doping practices. Urine samples were storage at −30 °C. Following instructions from the supplier, VAMS were dried for 2 h at room temperature. The dried materials were stored in sealed polyethylene bags containing silica gel at 4 °C until extraction.

Urine sample volume was fixed at 2.0 mL. Matrix-matched calibration curves were prepared by fortifying blank urine samples from healthy volunteers with isometheptene (2–250 ng/mL) or dorzolamide and brinzolamide (5–200 ng/mL). Deuterated internal standards (IS) (dorzolamide-d5, brinzolamide-d5, and isometheptene-d3) were added to all samples at a final concentration of 50 ng/mL.

For diuretics, liquid–liquid extraction was performed by adding 600 µL of 1 M ammonium formate buffer (pH 4.0) and 1 g NaCl. The pH was adjusted to 4.0 with glacial acetic acid when necessary. For isometheptene, alkaline liquid–liquid extraction was conducted using a Na₂CO₃/NaHCO₃ (1:3) buffer. Subsequently, 5 mL of tert-butyl methyl ether (TBME)/acetate buffer (1:1, v/v) was added, and samples were vortexed and centrifuged (3,000 rpm, 5 min). The organic layer was evaporated under nitrogen at 40 °C and reconstituted in 100 µL of mobile phase mixture (70:30 aqueous: methanolic). UHPLC-MS/MS analyzed urine extracts.

For calibration and quality control, venous whole blood collected in K₂EDTA tubes was fortified with isometheptene (5–50 ng/mL) or with dorzolamide/brinzolamide (5–200 ng/mL). Blank and fortified blood samples were applied to Mitra® VAMS tips and processed as described below.

For extraction, VAMS tip were detached and transferred to microtubes containing 5 μL of IS working solution contained either isometheptene-d3 or brinzolamide-d5/dorzolamide-d5 (50 ng/mL each) and 400 μL of extraction solvent (methanol/acetonitrile/2% aqueous acetic acid, 1:1:1, v/v/v). Samples were sonicated for 30 min at 30 °C, centrifuged (17,000 g, 5 min). Supernatants were evaporated under nitrogen at 40 °C, and residues were reconstituted in 100 µL of mobile phase (80:20, aqueous:methanolic) prior to UHPLC–MS/MS analysis.

Analyses were performed using a Dionex UHPLC system coupled to a TSQ Quantiva triple-quadrupole mass spectrometer (Thermo Fisher Scientific, Bremen, Germany) equipped with electrospray ionization source operating in positive ion mode. Chromatographic separation was achieved on a Syncronis C8 column (1.7 µm, 50 mm × 2.1 mm, Thermo Fisher Scientific) maintained at 40 °C. The mobile phases consisted of (A) water with 5 mM ammonium formate, (B) methanol, and (C) acetonitrile, all of which contained 0.1% formic acid. Injection volume was 10 µL.

For urine samples containing brinzolamide and isometheptene, the gradient program for mobile phase B was as follows: 15% (0 min) → 80% (17.5 min), held for 0.5 min, returned to 15% (18.1 min), and equilibrated until 20 min, with a flow rate of 0.3 mL/min. For dorzolamide urine samples, phase B started at 1%, increased to 5% (1.5 min), then to 90% (7 min), and was held for 2 min. It was then returned to 1% (9.1 min) and equilibrated for 15 min at a flow rate of 0.4 mL/min.

For VAMS samples, the gradient for phase C began at 10%, increased to 35% (25 min), then to 80% (30 min), and was held at 100% (32 min). It then returned to 10% (32.1 min) and equilibrated for 35 min at 0.3 mL/min.

The ion transfer tube and vaporizer were set to 350 °C and 400 °C, respectively. Spray voltage was static; sheath, auxiliary, and sweep gas flows were 40, 10, and 2 (arbitrary units). Selected reaction monitoring (SRM) transitions, collision energies, and precursor/product ions as follows: Isometheptene [M + H]⁺ m/z 142 → m/z 69 (16); m/z 111 (10); m/z 58 (13). Dorzolamide [M + H]⁺ m/z 325 → m/z 135 (30), m/z 151 (30), m/z 219 (30). Brinzolamide [M + H]⁺ m/z 384 → m/z 217 (30), m/z 163 (30), m/z 195 (30). For quantification purposes, the ISTD-normalized peak obtained from the extracted ion chromatograms of the first precursor/product ion pairs above mentioned in each analyte was adopted. Brinzolamide-d5 was monitored in urine using the mass transition [M + H]⁺ m/z 389 → m/z 136 (43). To improve specificity, the product ion was changed to m/z 281 due to an endogenous interference. Dorzolamide-d5 was monitored using transition [M + H]⁺ m/z 330 → m/z 135 (30), and isometheptene-d3 were monitored using the transition [M + H]⁺ m/z 145 → m/z 69 (15) independently of the matrix.

Peak width was set to 20 s, cycle time to 2 s, and Q1/Q3 resolution to 0.7 FWHM. Data acquisition and processing were performed using TraceFinder 4.2 software (Thermo Fisher Scientific).

Blank samples from different volunteers were used to build the set of replicates during the validation. The procedures used for semi-quantitative determination of the analytes in urine and capillary blood were validated following the analytical features:

Selectivity: Evaluated by analyzing drug-free urine and capillary whole-blood (VAMS) samples (n = 10 for each matrix) to verify the absence of endogenous or matrix-related interfering signals.

The methods for semi-quantitative analysis were validated and the results were tabulated in Supplementary Material 1.

The absence of interfering signals confirmed the specificity of the methods, and LOQs were 2–3 ng/mL for isometheptene, 4 ng/mL for dorzolamide, and 2–4 ng/mL for brinzolamide, depending on the matrix (urine or VAMS). Those values were established considering a double criterion: i) signal-to-noise ratio of 10 and ii) meeting identification criteria (39). Since matrix from different volunteer were adopted, the same values were considered the methods Limit of Identification (LOI). Linearity were acceptable having as criteria R² ≥ 0.99. Ion suppression or enhancement was observed (from 60% to 108%) depending on the analyte and matrix. The results were accepted considering the use of deuterated IS for each analyte. Intra-day and inter-day did not exceed 20% in the low concentration evaluated (5 ng/mL). For both parameters, imprecision was lower than 10% at 200 ng/mL. These results were considered acceptable considering no-threshold substances. The absence of detectable analyte signals in the blank samples after programed injection indicate absence of carry-over.

To ensure reliable concentration estimates, VAMS devices and isotope-labelled internal standards were employed were included to minimize potential matrix effects arising from interindividual variability in blood and urine composition. For urine, concentration values were adjusted to the specific gravity. The hematocrit level could potentially impact in the diffusion properties of the capillary blood in the card/paper sampling devices. Theoretically, the adoption of Mitra®/VAMS eliminates this impact (26, 40). Capillary and venous blood concentrations obtained at the first sampling time point were comparable, and no systematic bias attributable to hematocrit values (37.4–47.0%) was observed.

The stimulant class comprises hundreds of substances with wide structural and pharmacokinetic diversity (41). The subclass of alkylamines has drawn attention due to the high incidence of dietary supplement contamination. In Brazil, isometheptene is of special concern given its widespread availability in over-the-counter headache medications and the prevalent culture of self-medication (42). There are no reports in the literature on the evaluation of alkylamines by capillary blood. Furthermore, the class of stimulants is not included in the Testing Menu defined by dedicated DBS Technical (23). These factors motivated the selection of isometheptene as a model compound in an exploratory investigation.

Following a single oral dose of isometheptene, all volunteers presented detectable urinary concentrations at 2 h post-administration, corresponding to individual Cmax values and confirming rapid absorption. Marked interindividual variability was observed, with urinary Cmax values ranging from approximately 12,000 ng/mL to less than 800 ng/mL (Figure 1). In all volunteers, the concentrations in this collection point exceeded the MRL (50 ng/mL) by at least two orders of magnitude, clearly indicating an Adverse Analytical Finding (AAF), even though some values fell outside the validated linear range.

The urinary excretion profile demonstrated a rapid decline in isometheptene concentrations. In 70% of volunteers, levels fell below the MRL within 7 h of administration, indicating that therapeutic in-competition use could theoretically result in a negative finding. Volunteer 04 presented a urinary isometheptene concentration of 68 ng/mL at 24 h, and Volunteer 08 exhibited 89 ng/mL at 48 h post-dose, both cases corresponding to potential AAFs (Figure 1).

Although no MRL is currently defined for blood/capillary blood samples under WADA regulations, the comparison of matrices provides valuable insight. Isometheptene was detectable in capillary blood samples from all volunteers (2 h post-administration), mirroring the urinary data. As in urine, Cmax in capillary blood exhibited high interindividual variability, likely reflecting pharmacokinetic differences that are typically unavailable in anti-doping contexts. Therefore, only estimated concentrations were considered, in accordance with the routine protocol.

Isometheptene was undetectable in capillary blood samples collected 7 h' post-dose in all volunteers who also tested negative for isometheptene in urine at the same time window. In four cases, blood concentrations were detectable, but below the LOQ. Notably, 37% of urine samples that would produce an AAF had corresponding capillary blood concentrations below the LOQ.

Volunteers 04 and 08, whose urinary concentrations would trigger AAFs in a hypothetic out of competition period administration, illustrate this dissociation. For Volunteer 04, isometheptene was undetectable in capillary blood 24 h post-dose, with levels already below 5 ng/mL at 7 h. As expected, volunteer 08 showed a longer detectability of isometheptene in capillary blood. However, concentrations were approximately 7 ng/mL at 24 h and undetectable at 48 h after administration. These values are substantially lower than hypothetical DBS cut-offs proposed for other stimulants as cocaine (20 ng/mL), benzolecgonina (75 ng/mL) and amphetamine (20 ng/mL) in the literature (6), suggesting the suitability of these value ranges also for isometheptene.

The pronounced interindividual variability highlights the challenge faced by RMAs if only urinary MRLs were used as an interpretation tool and underscores the potential risk for athletes using isometheptene-containing medications. Similar concerns have been reported for other stimulants, including amphetamine, methylphenidate, and modafinil, which are also listed as doping agents (43). Based on these findings, a conservative washout period should be strongly recommended following administration.

The occurrence of AAFs based on urinary concentrations in the absence of detectable levels in capillary blood raises questions about the assertiveness of the MRL values established for urine. If the stimulant is not detected in blood, it likely indicates that the compound is either no longer pharmacologically active or has been rapidly eliminated. Consequently, urinary concentration alone may not accurately reflect the physiological impact of the substance at the time of sample collection, critical point considering in competition controls. These findings reinforce that urine-based MRLs have inherent limitations, and incorporating capillary blood through VAMS analysis could provide a more accurate basis for RMA (43).

Indeed, the shorter detection window observed in capillary blood samples suggests improved discrimination between in-competition and out-of-competition administration and may strengthen result adjudication in cases involving alleged inadvertent exposure, such as supplement contamination (11, 43, 44).

Considering the interindividual variability observed in urinary excretion, establishing a universal MRL for an entire stimulant class is particularly remains challenging, as similarly observed for glucocorticosteroids class (45). The detection or non-detection of a doping agent in capillary blood may provide complementary evidence for RMAs in doping cases involving substances prohibited in-competition only.

Finally, the structural and physicochemical diversity within the stimulant class underscores the importance of additional comparative studies. To our knowledge, this is the first investigation focused on the alkylamine subclass, providing new insights into its pharmacokinetic behavior and implications for anti-doping result interpretation.