Feasibility of dried blood spot collection for caffeine pharmacokinetic studies in microgravity: Insights from parabolic flight campaigns

Abstract

Aims

Therapeutic drug monitoring for astronauts faces limitations in conventional blood sampling and sample management onboard the international space station. Here, we explore the feasibility of dried blood spot (DBS) collection during parabolic flights (PF) to overcome these constraints.

Methods

We assessed the feasibility of blood deposition on blotting paper for preanalytical aspects in a PF using synthetic blood. Subsequently, DBS sampling validation was carried out in another PF campaign. Twenty volunteers participated in a pharmacokinetic study on caffeine and its metabolite, paraxanthine, conducted during parabolic flights. After >18 h caffeine washout, coffee (115 mg), tea (30 mg) or dark chocolate (11 mg) were ingested by the participants. DBS samples were collected at baseline, during weightlessness and post‐flight. Caffeine and paraxanthine were analysed using liquid chromatography–tandem mass spectrometry (LC‐MS/MS). Genotyping for cytochrome P450‐1A2 (CYP1A2) was performed and a metabolic ratio by area under the curve for caffeine and paraxanthine (AUC_PAX/AUC_CAF) for CYP1A2 was determined. A user experience survey was also conducted.

Results

Full in‐flight pharmacokinetic study was feasible in seventeen volunteers with three unable to perform the sampling due to motion sickness. Nineteen participants (twelve males and seven females) completed pharmacokinetic profiles, with repeated pharmacokinetic studies for six participants. CYP1A2 genotyping resulted in eight ultrarapid, eleven intermediate, and one poor metabolizer. Among the women, four were on oestrogen contraceptives, a known inhibitor of CYP1A2, and were considered as poor metabolizers. Expected differences in kinetic profiles, consistent with consumption habits, the ingested dose and the genotypic/phenotypic information, were observed. The mean caffeine AUC for coffee, tea and chocolate were 9419 ng.h.mL ⁻¹ (95% confidence interval [CI]: 6222–12 616, n = 10), 6917 ng.h.mL ⁻¹ (95% CI: 2729–11 105), n = 7) and 3039 ng.h.mL ⁻¹ (95% CI: 1614–4142, n = 12), respectively. The mean paraxanthine AUC were 10 566 ng.h.mL ⁻¹ (95% CI: 6242–14 890, n = 10), 4011 ng.h.mL ⁻¹ (95% CI: 2305–5716, n = 7) and 3638 ng.h.mL ⁻¹ (95% CI: 1589–40 859, n = 12), respectively. The mean metabolic ratio in oestrogen‐treated women was 0.53 (95% CI: 0.35–0.71) compared to 1.19 (95% CI: 0.99–1.33) in others. The mean metabolic ratio was 1.02 (95% CI: 0.81–1.23, n = 15) on the ground and 1.10 (95% CI: 0.70–1.41, n = 13) during the parabolic flights, with no significant difference observed between the two conditions. Overall participants were satisfied with the usability of the method.

Conclusions

DBS collection was safe, stable, feasible and well accepted in weightlessness. This method would offer valuable insights into human metabolism adaptation during long‐term spaceflight, addressing space pharmacology challenges.

What is already known about this subject

• Therapeutic drug monitoring (TDM) in space lacks established methods due to sample management challenges and space constraints.

• Microsampling techniques like dried blood spot (DBS) have shown promise in terrestrial settings.

• DBS has never been validated for pharmacology use in microgravity conditions.

What this study adds

• DBS microsampling is feasible in microgravity conditions, as demonstrated through a parabolic flight study.

• Caffeine pharmacokinetics assessed via DBS in microgravity conditions showed agreement with ground‐based studies.

• CYP1A2 phenotyping using DBS offers potential for monitoring drug metabolism in astronauts and could be extended to other metabolism pathways.

1. INTRODUCTION

Therapeutic drug monitoring (TDM) studies in space have faced challenges over the past 50 years of space exploration due to issues related to sample management, including venepuncture, blood processing, sample degradation, and storage constraints, microgravity and limited crew time, imposed by space conditions. Because of these constraints, research on drug metabolism under microgravity is limited, with few studies conducted during parabolic flights ¹ , ² , ³ , ⁴ , ⁵ , ⁶ and/or spaceflight. ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³ Factors such as cytochrome activity, glucuronidation enzymes and transporters, influenced by genetics and dynamic mechanisms such as drug interactions, impact pharmacokinetic variability. The absence of knowledge on the fate of drugs in spaceflight jeopardizes space missions and disease management in space. ¹⁴

Dried blood spot (DBS) sampling involves collecting capillary blood onto an absorbent paper card after performing a finger prick. The card is then allowed to dry, making it fast, minimally invasive and suitable for self‐collection by patients. DBS has been extensively used since the 1960s, particularly in neonatal disease screening, for example in Guthrie tests and many other applications, including TDM. ¹⁵ It facilitates sample storage, transport, stability, and may be advantageous for medication monitoring, especially during space missions where drug quantification can be determined retrospectively upon the samples' return to Earth. DBS is suitable for repeated sampling allowing for a more precise description of drug pharmacokinetic parameters. These advantages have led to the development of various applications in toxicokinetic studies, preclinical or clinical pharmacokinetic studies and therapeutic drug monitoring, ¹⁶ , ¹⁷ , ¹⁸ notably for determining area under the curve (AUC) of drug concentrations. In pharmacokinetic studies, AUC is a good index of exposure to drugs which could help in the prediction of therapeutic effect (toxicity or efficacy).

We set up the hypothesis that DBS could be used to fulfil pharmacokinetic studies during space missions. All steps, from blood collection to analyte quantification need to be validated in realistic conditions. Caffeine was used as a proof of concept, because of its well‐understood pharmacokinetics and minimal protein binding characteristics. ¹⁹ Caffeine is also a commonly used probe substrate for cytochrome P450 CYP1A2 phenotyping, an enzyme involved in drug metabolism and exhibits significant interindividual variability ¹⁹ of genetic, non‐genetic and environmental origin. We also relied on parabolic flight (PF) as it mimics several aspects encountered during space missions, including microgravity, confinement and mental stress. Our primary objective was to test the feasibility of depositing blood containing caffeine (CAF) and its metabolite paraxanthine (PAX) onto paper cards, at different haematocrit levels and concentrations for synthetic blood, and after exposure to alimentary caffeine in healthy volunteers under microgravity conditions in two successive PF campaigns. The secondary objective was to test the influence of CYP1A2 genotype and phenotype including caffeine‐use habits. The tertiary objective was the acceptance of the whole procedure by the volunteers.

2. METHODS

2.1. Bioanalytical methods to quantify caffeine and paraxanthine from DBS

Caffeine and paraxanthine dosage methods were validated in the lab following in‐house protocols in accordance with EMEA (European Medicines Agency) for validation of classical parameters (EMEA/CHMP/EWP/192217/2009, available at: https://www.ema.europa.eu/en/documents/scientific-guideline/guideline-bioanalytical-method-validation_en.pdf. DBS‐specific parameters such as haematocrit effect were evaluated in accordance with the International Association of Therapeutic Drug Monitoring and Clinical Toxicology (IATDMCT) guidelines ²⁰ (Supplemental data S1).

Validation was done twice because of change in equipment between the two PF campaigns. In brief, caffeine and paraxanthine were added to synthetic blood (XN CHECK®, SYSMEX) to obtain eight levels of concentration (from 25 [1st PF campaign] or from 50 [2nd PF campaign] to 5000 ng.mL⁻¹). A 20 μL spot volume of synthetic blood deposited onto the DBS paper card (Whatman® 903 protein saver) was selected for the validation. Concentrations of caffeine, paraxanthine and the internal standard (IS) d9‐caffeine, within the same DBS samples, were determined using a validated liquid chromatography triple quadrupole mass spectrometry method (Xevo TQD, Waters) operating in a positive electrospray ionization mode and targeted multiple reaction monitoring (MRM). Data acquisition and processing were carried out with the Mass‐Lynx® 4.1 software.

2.1.1. Method for the 1st PF campaign

The LLOQ was 25 ng.mL⁻¹ corresponding to the lowest level of calibration (Table S2.1). The results of our first validated quantification method for caffeine and its metabolite paraxanthine are shown in Table S2.2. Standard curves were linear (correlation coefficients r² > 0.995). Accuracy and precision were calculated on two levels of quality control (QC): LQC (low QC) = 125 ng.mL⁻¹, HQC (high HQC) = 1250 ng.mL⁻¹. Within‐ and between‐day precision (expressed as coefficient of variation, CV in %) and accuracies assessed on the two levels of QC are provided in Table S2.3A and S2.3B. For each analyte, the worst intraday inaccuracy observed was −17.7% for caffeine and 19.4% for paraxanthine. The highest CV observed for the intraday precision was 10.1% for caffeine and paraxanthine.

2.1.2. Method for the 2nd PF campaign

The LLOQ was 50 ng.mL⁻¹ corresponding to the lowest level of calibration (Table S3.1). The results of our second validated quantification method for caffeine and its metabolite paraxanthine are shown in Table S3.2. Standard curves were linear (correlation coefficients r² > 0.995). Accuracy and precision were calculated on three levels of quality control (QC): LQC (low QC) = 100 ng.mL⁻¹, MQC (mid QC) = 1000 ng.mL⁻¹, HQC (high HQC) = 2550 ng.mL⁻¹). Within‐ and between‐day precisions (expressed as coefficient of variation, CV in %) and accuracies assessed on the three levels of QC are provided Table S3.3A and S3.3B. For each analyte, the worst intraday inaccuracy observed was −21.5% for caffeine and 21.9% for paraxanthine. The highest CV observed for the intraday precision was 23.5% for caffeine and 23.1% for paraxanthine.

For detailed methods and results for the 1st and 2nd PF campaigns, see Supplemental data S2 and S3, respectively.

2.2. Parabolic flight: generalities

Parabolic flights were performed with a modified A310 Zero‐G airplane that follows specific flight manoeuvres aiming to produce 22 s of near weightlessness, close to “0 g”, every parabola. Each of those starts with a pull‐up and ends with a pull‐out: both of these phases last 20 s and generate hypergravity at 1.8 g (Figure S4). A typical flight consists of 30 subsequent parabolic manoeuvres, and most campaigns include three flight days in a row (https://www.airzerog.com/). PF was performed in March 2022 (synthetic blood) and March 2023 (caffeine clinical trial). Study designs and flow charts are presented in Figure 1A and B.

FIGURE 1.

A. Study design and flow chart for the 1st parabolic flight campaign. B. Study design and flow chart for the 2nd parabolic flight campaign.

2.3. DBS feasibility test in weightlessness

The main objective of the first parabolic flight campaign (VP164/62) was to investigate potential anomalies generated by microgravity in the passive absorption of capillary blood by paper cards. Additionally, the study aimed to assess the performance of the pipette for blood microsampling, evaluate the impact of PF conditions on CAF/PAX stability during DBS card drying, and determine the efficiency of blood‐related operations in microgravity (sampling and deposition).

We tested the deposition of synthetic blood (XN CHECK®, SYSMEX), 20 μL spot volume onto the DBS paper card (Whatman® 903 protein saver) overloaded with known concentrations of caffeine and paraxanthine at two haematocrit levels per the manufacturer (“normal” and “high”, 33.1% and 42.3%, respectively) during PF microgravity:

• 1 blank (G0, absence of caffeine and paraxanthine),

• 3 points both haematocrit level blood (G1, G3, G5): 25 ng.mL⁻¹; 75 ng.mL⁻¹; 500 ng.mL⁻¹

A standard air‐displacement pipette (FINNPIPETTE F1) was used for blood deposition. This type of pipette has already been employed and validated in microgravity conditions. ²¹

The experiment was conducted within a rack transformed into a glove box to confine the samples, reproducing a pharmacology laboratory bench. Operators were categorized into three levels of lab expertise: expert, intermediate, and “beginner” (n = 4 operators). Each operator performed the deposition at each concentration on DBS samples onboard the aircraft the day before the flight, allowing for comparison with in‐flight samples (Figure 1A).

Two Eppendorf (1.5 mL) tubes for each concentration and haematocrit were prefilled on the ground. The operator collected during the 20 s of normogravity and deposited during the 22 s of microgravity. Each card was attached to a clamp in the rack during the entire flight, thus being allowed to dry. Once completely dry, the card was then placed in a plastic bag and stored at room temperature until analysis (3 months after the campaign).

2.4. Parabolic flight campaign and DBS validation in a caffeine pharmacokinetic study

2.4.1. Participants

The study was approved by the national Ethics Committee (2022‐A02794–39), and declared as NCT06431984. We included a total of 20 healthy volunteers including 15 volunteers already scheduled for PF on other scientific experiments not involving drug administration, and five members of the team. All participants had flight medical approval, signed informed consent forms and were subjected to an on‐site clinical check of their health status. Experiments were conducted during a PF campaign (VP171/64).

In addition to demographics, the following parameters were assessed: medications metabolized by CYP1A2 (including the use of oestrogenic contraceptives), caffeine consumption frequency categorized into three groups: high (>3 cups of coffee or tea per day), moderate (2 cups of coffee or tea per day) and low (1 cup of coffee or tea or occasional). Genotyping of CYP1A2 on saliva samples was performed, the samples having been collected on site, prior to the flights.

2.4.2. Study design

After an 18 h caffeine washout, participants were given, according to personal preference, either coffee (115 mg of caffeine) or tea (30 mg), or three dark chocolate squares (11 mg). All participants had a 30 min training session under supervision by our team, then all DBS were done in total autonomy. Four to five blood samples were auto‐collected at different times at baseline on the ground (T0), during microgravity 2 h to 4 h post ingestion (T2), and twice post‐return on the ground 5 h to 6 h post ingestion (T3, T4). Four participants repeated the study, conducting DBS sampling during the 3 days of PF, while two participants repeated on two consecutive days (Figure 1B).

Five volunteers of the investigation team repeated the study on the ground for each caffeine presentation after 72 h of caffeine washout (12 DBS over 12 h, 3 times).

2.4.3. DBS quality analysis and caffeine quantification

For safety issues, samples were collected in a closed rack converted into a glove box, thus preventing any potential blood droplets from free‐floating into the cabin. The rack was disinfected between use by each volunteer. Each card offered five preprinted circles of standard size (12 mm diameter) for guiding collection. Cards were considered valid for analysis if spots were large enough to allow for a single small (3 mm) hole punch, which means that at least one drop of blood in one circle is necessary to analyse the card. Volunteers were encouraged to try and deposit on each of the five circles, if possible. Any obviously smeared or double dropped spots were not considered valid. Caffeine and paraxanthine were analysed using liquid chromatography–tandem mass spectrometry (LC‐MS/MS) after 6 months' storage at room temperature, mimicking a potential spaceflight sample return timeline.

2.4.4. Genotyping

Saliva samples were collected using ORAGEN®‐DNA saliva collection kit. DNA was extracted from the samples using the Qiagen technique on the Qiasymphony automated system. The participants were genotyped for different polymorphisms: CYP1A2 (CYP1A2*1F C734A rs762551) using the TaqMan allelic discrimination assay. The predicted phenotypes were based on enzyme activity of the identified alleles, as listed in PharmGKB (https://www.pharmagkb.org/).

Subjects were categorized for CYP1A2 into one of three possible genotypes: A/A ultrarapid metabolizers (UM), C/A intermediate metabolizers (IM) or C/C poor metabolizers (PM).

2.4.5. Phenotyping

Phenotype determination was based on the metabolite to parent drug metabolic ratio (MR). As for genotyping, participants were classified as PM_phe, IM_phe and UM_phe according to their MR.

Area under the curve from time 0 to last for both caffeine and paraxanthine (AUC_0‐Last,CAF and AUC_0‐Last,PAX, respectively) was calculated using the trapezoidal method with linear interpolation (Microsoft Office Excel version 16.83), only for participants who had at least three sampling times. Caffeine metabolic ratios based on AUC_0‐Last were calculated using molar concentrations.

2.4.6. End‐user experience survey

Post‐flight paper questionnaires were submitted to volunteers to collect feedback, evaluating participant satisfaction and ease‐of‐use of DBS sampling, each question using a 5‐point Likert scale (Supplemental data S5). Study management and research usefulness for space exploration was also covered, as well as the quality of provided information and instructions, participant apprehension of microgravity, DBS practicality and in‐flight performance. Participants were also asked about their views on potential self‐sampling for future space missions and were encouraged to provide insight for future methodology improvement. None of the research team members participated in the feedback collection process.

2.4.7. Data analysis

The feasibility of DBS sampling was considered conclusive if 90% of ground samples and 80% of in‐flight samples were analysed. Statistical analysis was performed using the following software: NCSS (version 2022) and GraphPad Prism (version 10.2.0 [335] 2024). The General Linear Model (GLM) was employed for analysing the results of the first PF campaign, with statistical significance determined at P < .05. In the second PF campaign, statistical analyses were performed using a one‐way ANOVA test, followed by Tukey post hoc test to allow for multiple comparisons (CYP1A2 phenotyping, participant's flight vs. ground). Additionally, an unpaired t‐test was used to compare the MR according to caffeine consumption. Data analysis included descriptive statistics for closed‐ended questions and thematic/content analysis for open‐ended questions.

3. RESULTS

3.1. First parabolic flight campaign

3.1.1. DBS feasibility in weightlessness

Blood transfer in microgravity proved to be feasible with the tested pipette (Figure 2). Consistency between in‐flight and on‐ground concentration deposits at both haematocrit levels within the aircraft was revealed during analysis, indicating no significant impact of microgravity on the deposition process (Figure S6). Stored samples remained stable for three 3 months after the campaign, exhibiting no variance in concentrations compared to theoretical levels. However, at 500 ng.mL⁻¹, significant differences were noted for caffeine and paraxanthine compared to theoretical laboratory concentrations (Figure S7). Operator experience did not influence sample handling (Figure S8) or resultant concentrations. Additionally, no significant difference was observed between caffeine and paraxanthine concentrations measured in discs punched out peripherally vs. centrally, irrespective of the haematocrit level, consistent with laboratory test assays (Table S4).

FIGURE 2.

Feasibility of blood deposition with a pipette during weightlessness.

3.2. Second parabolic flight campaign: human application

3.2.1. Subjects

A total of 20 participants were enrolled, thirteen men and seven women, aged 38.1 ± 12.5 years. They were all non‐smokers, with an average body mass index of 24.1 ± 2.6 kg/m². Four of the women reported using oestrogenic contraceptives. Additionally, eight participants reported high caffeine intake, six reported moderate intake, and the other six reported low intake.

3.2.2. DBS sample quality

Out of the 20 volunteers, 17 successfully deposited blood drops from their fingers in microgravity within the allotted 22 s. The blood sampling (finger prick) was conducted during the 1.8 g phase, while the capillary collection (blood drop deposition onto the paper card) took place during the 0 g phase. Among the participants, six of them were able to perform the finger prick and the blood deposition within the 0 g phase, in 22 s (Figures 3 and S9). Three volunteers were unable to perform the procedure due to severe motion sickness. All of 128 collected DBS cards (n = 87 on the ground, n = 41 in weightlessness) were considered suitable for the quantification of caffeine and paraxanthine. The median number of deposits in weightlessness was three and 24.4% of participants reached the maximum of five deposits (Table 1). In contrast, on the ground, the median number is four (64% reaching five spots). When feasible, two spots were analysed for each card at every time point in order to ensure repeatability. A total of 88.7% of dried blood spots deposited in flight were considered valid for analysis, whereas 96.7% of deposits performed on the ground were validated. A summary of DBS card quality and the quantity of blood spots collected by each participant on the ground vs. in weightlessness is presented in Table 1.

FIGURE 3.

Dried blood spot collection during weightlessness. A: finger prick test; B: blood spotting on the paper card.

TABLE 1.

Summary of DBS card quality and the quantity of blood spots collected during the second parabolic flight campaign.

Card quality (number of analysable spots)	% of total cards on the ground	% of total cards in weightlessness
1 spot	0	9.8
2 spots	8.0	34.1
3 spots	13.8	22.0
4 spots	13.8	12.2
5 spots	64.4	24.4
Total spots, n = 537	386	141
Total analysable spots, n = 508	383	125
% of total analysable spots	96.7	88.7

3.2.3. Caffeine pharmacokinetic results

A total of 19 of the 20 volunteers (12 men and 7 women) completed pharmacokinetic profiles, with six subjects repeating the pharmacokinetic study. The mean and standard deviation caffeine AUC for coffee, tea and chocolate were 9419 ± 393 ng.h.mL ⁻¹ (n = 10), 6917 ± 278 ng.h.mL ⁻¹ (n = 7), 3039 ± 203 ng.h.mL ⁻¹ (n = 12), respectively, and the mean paraxanthine AUC were 10 566 ± 400 ng.h.mL ⁻¹ (n = 10), 4011 ± 104 ng.h.mL ⁻¹ (n = 7), 3638 ± 150 ng.h.mL ⁻¹ (n = 12), respectively. We observed a direct proportionality between ingested dose and caffeine exposure during the parabolic flight campaign and on the ground (Figures 4, S10A, S10B and 6). The baseline for chronic consumers (moderate to high) was above the limit of quantification. The calculated AUCs allowed us to discriminate the three distinct caffeine consumer profiles (high, moderate, low) (Figures 5, S11A and S11B). Caffeine and paraxanthine exposure variability on the ground for five team members from our team, according to the caffeine ingested form, is shown in Figure 6, and supplemental Figures S12A and S12B.

FIGURE 4.

Caffeine and paraxanthine exposure according to caffeine ingested form (coffee, tea, chocolate).

FIGURE 6.

Caffeine and paraxanthine exposure on the ground for one participant (coffee, tea, chocolate).

FIGURE 5.

Caffeine and paraxanthine exposure according to caffeine consumption frequency (low, moderate and high).

For CYP1A2 genotyping, eight subjects had the A/A genotype (45%), whereas eleven subjects had the C/A genotypes (50%). One participant was homozygous (C/C) but was excluded due to having only two sampling times. Among the seven women, four were on oestrogen contraceptives (n = 3 A/A and n = 1 C/A), a known inhibitor of CYP1A2, and were thus considered as PM.

The distribution of phenotype ratios within each genotype is shown in Figure 7. When comparing individual genotyping results with phenotyping, there was 80% (16/20) overall concordance between genotype and phenotype. The 20% of discordance can be explained by the inclusion of women using oestrogenic contraceptives. The MR CYP1A2 in oestrogen‐treated women was 0.53 ± 0.17 compared with 1.19 ± 0.36 in others, confirming their PM status. Participants who had higher caffeine intake (3 cups of coffee or tea/day) showed a non‐significant trend for higher CYP1A2 activity (MR) than participants with lower average caffeine consumption (P‐value = .07) (Figure S13). Furthermore, the MR did not significantly differ between parabolic flight conditions and ground conditions for the five subjects (P > .05, paired t‐test), regardless of any form of caffeine (Figure 8). The within‐participant coefficient of variation for MR was 16.11 ± 9.16 (n = 6, data not shown).

FIGURE 7.

Metabolic ratios of caffeine according different phenotype.

FIGURE 8.

CYP1A2 MR comparison between parabolic flight conditions and the ground for five participants for each form (coffee, tea, chocolate).

3.2.4. End‐user experience survey

Fourteen out of 15 (93%) volunteers completed the questionnaire, rating each question on a scale of 1 to 5. The five members of the research team have been excluded from the feedback collection process. Overall, the volunteers showed a high level of satisfaction with usability (mean = 4.9, median = 5). Specifically, 86% expressed complete satisfaction with the management of the research and 14% rated it as satisfactory. The volunteers were also highly satisfied with various aspects of the research, including the level of information provided to the participant, communication quality with the research team, DBS information quality and the DBS sampling process (means = 4.9, medians = 5.0). Detailed results are presented in Table S5. Regarding sampling experience, 79% reported a lack of worry or anxiety, with only 21% experiencing mild anxiety. All participants (100%) expressed a positive view of the self‐sampling procedures in the context of future space missions.

Participant qualitative feedback highlighted several points:

• Interest in DBS use during space missions: Participants showed positive interest in using the DBS method, describing it as “innovative”, “interesting” and “useful”.

• Challenges and recommendations: Some participants noted difficulties in blood extraction, suggesting alternative sites, larger lancets, and pre‐marking the finger for better precision.

• Method adaptability in microgravity: Suggestions for improvement included using a more ergonomic stylus‐type device for self‐sampling. Additionally, participants found the method to be “very reliable”, “accessible”, “feasible during missions in total autonomy”, “simple” and “effective”.

4. DISCUSSION

The main findings of the present study were that blood drop transfers on paper cards were feasible in microgravity, and that volunteers were able to self‐collect samples. Another finding is that with the proposed methodology, repeatable kinetic profiles—consistent with consumption habits, the ingested dose and the genotypic/phenotypic information—were obtained. The feasibility of both collection and analysis of DBS is indicative of usability and potential success in future space pharmacology studies. Despite astronauts being selected for their excellent physical and mental health, they often require medication during missions for various reasons, including treatment of space adaption syndrome, urinary tract infections or venous thrombosis. ¹⁴ , ²² , ²³ Additionally, medications may be used preventively to counteract space‐induced vascular ageing, bone demineralization, fat redistribution or mitigate radiation effects. ²⁴ Since very little is known about pharmacokinetic and pharmacodynamic (PK/PD) in space conditions, the proposed method using DBS sampling could help to fill the gap.

4.1. Feasibility

The development of reliable and safe methods for collecting biologicals fluids during spaceflight is crucial for the development of TDM during space missions, ensuring crew health and ultimately their success. The feasibility of DBS sampling as an alternative to venepuncture during parabolic flight campaigns, using caffeine and paraxanthine exposure, was tested. In the present study, the quality of the samples was satisfactory. As suggested by some participants, DBS quality could be improved by using larger needle lancets to achieve a sufficient spot size.

Droplets behave differently in microgravity: they are spherical and only adhere to supports by surface tension. ²⁵ During the first parabolic flight campaign, it was observed that the drop of blood adheres to the pipette and stays in place during microgravity. Once in contact with the paper card, the large absorbent surface retained the drop, a complete transfer being achieved without any free‐floating elements. For finger prick drops tested during the second flight campaign, drops of blood tended to adhere to the finger without collapsing and were spherical. This positive result was not self‐evident since previous experiments showed that arterial and venous bleeders broke into uniform, low‐velocity spheres which bounce off absorbent pads and suction tips; conventional dabbing with gauze tending to create small blood spheres. ²⁶ The surface of the selected blotting paper card appears adequate to absorb blood drops quickly and completely. We did not experience prolonged bleeding; however, microgravity was limited to 22 s periods and adhesive band compression was applied immediately after blood collection. One potential risk is that the drop of blood becomes too large to adhere to the finger and starts to free‐float, which led us to enclose the experiment in a glove box. Although this did not happen for 125 drop transfers, it is still reasonable to confine the blood collection procedure. Drying the cards in the glove box did not appear an issue, as the 20–40% humidity rate in the cabin, coupled to the regular opening of the box for cleaning, allowed for easy drying.

4.2. Reliability

Quantification of caffeine and paraxanthine in non‐volumetrically applied capillary DBS differs from venous blood and plasma‐based quantification. ²⁷ Analytically, considering punch localization, blood volume and haematocrit is crucial. Variations in blood volume spotted did not impact significantly caffeine and paraxanthine concentrations. However, we observed a trend of increasing concentrations with higher blood volumes, consistent with results reported by De Kesel et al. ²⁷ Despite this trend, differences from normalized samples remained within 15%. Seemingly, microgravity does not influence blood spreading in DBS, resulting in uniform analyte distribution. It might eventually work better than on the ground since only surface tension forces apply in the plane of paper, not gravity. ²⁸ Additionally, haematocrit affects caffeine and paraxanthine quantification in DBS, with higher concentrations measured at high haematocrit and lower concentrations at low haematocrit. However, their molar ratio (PAX/CAF) mitigates these issues, ²⁷ , ²⁹ , ³⁰ , ³¹ , ³² supporting the utility of DBS‐based phenotyping.

Recent astronaut data showed significant haematocrit increases during spaceflight. As the haematocrit level can potentially be ascertainable on board the ISS, a correction factor can be applied. Alternatively, other volumetric microsampling strategies ³³ , ³⁴ , ³⁵ could be used to mitigate this bias. ³⁶ For the present study, Neoteryx Mitra® devices were discarded due to their higher weight and cost compared to blotting papers, as well as the risk of under‐ or oversaturation of the tips in suboptimal conditions. Additionally, the DBS method allows for the punching of multiple spots on the paper card, providing flexibility for further analysis and biobanking.

Some studies have explored the feasibility of using a single blood sample rather than relying on the AUC ratio to precisely evaluate CYP phenotyping indices. ³⁷ These studies have indicated that the optimal sampling time for CYP1A2 is at T3 h. For its application in space pharmacology, since the central issue is to establish a cartography of the main pharmacokinetic parameter changes, the AUC ratio approach appears to be the most pragmatic.

Stability tests were not conducted under extreme temperatures or over extended periods in the present study. We recently published the good stability of DBS/DUS under extreme conditions. ⁵⁷ De Kesel et al., ²⁹ showed that DBS remained stable up to 324 days at room temperature and 4 days at 50 °C. This stability is advantageous for transporting samples from the ISS to Earth for analysis. Further tests aiming to assess the potential impact of radiation on DBS must be conducted aboard the ISS or in low Earth orbit.

4.3. Relevance

While genotype is immutable in a patient, it does not predict protein expression and functionality. As evidenced in our study, women using oestrogen contraceptives had doubling AUC and MR, irrespective of their genotype. The interest in pharmacological phenotyping (in addition to genotyping) will be to offer prediction of drug efficacy, toxicity and interactions, thus providing truly personalized drug therapy in astronauts.

Similar CYP1A2 activity between parabolic flight and ground conditions can likely be attributed to short exposure time to altered gravity; it does not, however, preclude changes during longer space missions. Spaceflight induces important physiological changes: 30% reduction in muscle mass, impaired glucose and lipid metabolism and a low‐grade inflammation immune system. ³⁸ , ³⁹ , ⁴⁰ Inflammation significantly alters liver gene expression profiles, often leading to the downregulation of drug metabolizing enzymes. ⁴¹ , ⁴² , ⁴³ These effects can partly be attributed to the interplay of inflammatory signalling pathways and transcriptional suppression, potentially resulting in variations in drug pharmacokinetics and therapeutic effects. ⁴⁴ , ⁴⁵

Furthermore, astronauts can develop a diabetogenic phenotype linked to flight duration. ⁴⁶ Additionally, bed rest, an earth‐bound simulated microgravity, induces non‐alcoholic fatty liver disease (NAFLD), ⁴⁷ characterized by abnormal intrahepatic fat accumulation. NAFLD, prevalent in sedentary individuals, is associated with metabolic dysregulation and insulin resistance. The enzymatic activity of CYP3A4 decreases with the severity of NAFLD in type 2 diabetic patients and may be related to inflammatory processes. ⁴⁸ , ⁴⁹ , ⁵⁰ , ⁵¹ Considering the metabolic stress observed during space missions ⁵² or bed rest studies, ⁵³ the development of NAFLD in astronauts may influence CYP activity, thus contributing to drug variability. ⁵²

4.4. Perspectives

Expected improvements in DBS, including automation, ³³ , ⁵⁴ , ⁵⁵ make it an interesting method for dosages in harsh milieus including space, eliminating storage needs and enabling on‐site analysis. This outlook on new blood analysis methods would benefit not only Earth‐based medical teams, but also space research.

Other biological fluids which can be collected by non‐invasive means, such as saliva and urine, could be considered for TDM. However, the composition of saliva may change during flights, as indicated by findings in mice, ⁵⁶ and this could introduce additional variability to drug dosing if such a quantitative approach were adopted for astronauts. Urine, on the other hand, is an interesting matrix for TDM microsampling and can accurately reflect drug elimination mainly by the kidneys. We demonstrated the feasibility of dried urine spot sampling in microgravity during a PF for detecting potential cardiovascular drug candidates aimed at preventing early vascular ageing in astronauts. ⁵⁷

In the list of the 78 drugs permanently available on the ISS (year 2014), ⁵⁸ 24 (31%) are affected by genetic polymorphisms. ⁵⁹ In neuropsychopharmacology ⁶⁰ TDM: one is strongly recommended (phenytoin), three are recommended (venlafaxine, aripiprazole, sertraline), one is useful (modafinil) and three are potentially useful (diazepam, lorazepam, promethazine). For the anti‐infective drugs class: four antibiotics (amoxicillin, clindamycin, levofloxacin, sulfamethoxazole), one antifungal (fluconazole), and one antiviral (valacyclovir) are typically monitored in hospital settings for critically ill patients and outpatients with long‐term conditions. ⁶¹ , ⁶² , ⁶³ , ⁶⁴ , ⁶⁵ For several drugs, DBS is already used on Earth for TDM purposes. ⁶⁶ , ⁶⁷ , ⁶⁸

4.5. Limitations of the study

The washout period proved to be insufficient during the parabolic flight study. A 72‐h washout duration would have been optimal, given the interindividual variability of caffeine and its half‐life ranging between 2.3 h and 9.9 h. ¹⁹ However, implementing this washout period was not feasible due to the flight campaign schedule and the inability to monitor every volunteer. Nonetheless, a 72‐h washout duration was conducted for participants who repeated the kinetics on the ground, with residual concentrations remaining below the limit of quantification. The study was designed for feasibility under microgravity, not for relevance of findings related to the effect of microgravity. Additionally, while this study demonstrated the feasibility of using DBS sampling in microgravity, it does not include a comparison with traditional blood sampling methods, such as venepuncture. This comparison was not feasible during the study due to the constraints of the PF environment. Future studies could include traditional blood sampling as a reference point for longer spaceflight missions.

Another limitation is that the study's sampling schedule, because of time constrains, only included four time points and no early point. This did not allow the identification of the true maximum concentration (C _max) of caffeine and paraxanthine, leading to an underestimation of peak exposure levels. On the ground study, we demonstrated that an early peak is detectable. Increasing the sampling frequency to more accurately characterize pharmacokinetic profiles and better reach C _max should be considered in future research, particularly in the context of extended space missions. For spaceflight studies, it would be preferable to wipe off the first drop to avoid potential interstitial fluid contaminations.

It is important to acknowledge that the precision of the method employed in this study did not fully meet standard bioanalytical requirements. However, this limitation does not impact our primary conclusions, as the main goal was to assess the feasibility of the entire procedure, from pipette deposits to blood collection and LC‐ MS/MS analysis. Despite only moderate accuracy, our method was adequate for highlighting the effects of genotype, caffeine consumption and oestrogen inhibition of CYP1A2 enzyme. Therefore, while the imprecision may have affected our ability to detect smaller differences, it does not alter our conclusions about the overall feasibility of the procedure which allowed us to detect the usual cause of interindividual variability in caffeine pharmacokinetics observed in clinical studies. Moreover, the protocol also highlighted a trough caffeine concentration depending on caffeine consumer profiles.

Although we relied on independent volunteers from other research teams and excluded our own team members to minimize bias, some degree of positive bias may still be present. Participants who were already involved in other experiments aboard the aircraft might have given more favourable feedback due to their awareness of the study's goals.

5. CONCLUSION

This study has demonstrated the feasibility of DBS in microgravity, thus offering a microsampling approach to TDM for astronauts. These results pave the way for future research on individualized preventative pharmacotherapy in space, and wider use of PK/PD for earth applications.

AUTHOR CONTRIBUTIONS

Audrey Derobertmasure: Conceptualization; investigation; validation; writing—original draft; visualization. Li Shean Toh: Investigation; formal analysis; English proof reading. Céline Verstuyft: Investigation; formal analysis. Srboljub Lukic: Formal analysis. Christelle De Sousa Carvalho: Investigation; validation; formal analysis. Raphaël Couronné: Investigation; formal analysis. Marie Beauvalet: Investigation; English proof reading. Pierre Boutouyrie: Conceptualization; investigation; writing—original draft; validation; supervision. Stéphanie Chhun: Conceptualization; investigation; writing—original draft; validation; supervision. All authors gave final approval for the manuscript.

CONFLICT OF INTEREST STATEMENT

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Supporting information

Data S1: Volume and haematocrit effect assays.

Figure S2.1. Boxplots representing haematocrit effect on caffeine and paraxanthine concentrations at high concentrations.

Figure S2.2. Boxplots representing volume effect on caffeine and paraxanthine concentrations at low and high concentrations.

Data S2: DBS method for the 1st parabolic flight campaign: calibration curves and LLOQ, accuracy and precision.

Table S2.1. The lowest quantification limits of analysed caffeine and paraxanthine in blood obtained by a dried blood spot.

Table S2.2. Average parameters of the linear regression obtained from three calibration curves of the analytes.

Table S2.3A. Within‐day precision and accuracy for the two quality control levels and for each validation day.

Table S2.3B. Between‐day precision and accuracy on two quality controls.

Data S3: DBS method for the 2nd parabolic flight campaign: calibration curves and LLOQ, accuracy and precision.

Table S3.1. The lowest quantification limits of analysed caffeine and paraxanthine in blood obtained by a dried blood spot.

Table S3.2. Average parameters of the linear regression obtained from three calibration curves of the analytes.

Table S3.3A. Within‐day precision and accuracy for the two quality control levels and for each validation day.

Table S3.3B. Between‐day precision and accuracy on two quality controls.

Figure S4. How do parabolic flights work?

Data S5: Survey to the participants of the parabolic flight campaign.

Figure S6. Effect of haematocrit level (a) and weightlessness DBS sampling (b) on the mean residual errors of caffeine/paraxanthine concentrations (difference between measured concentration and theoretical concentration). H (high haematocrit), N (normal haematocrit).

Figure S7. Analysis of mean residual errors of caffeine (a) and paraxanthine (b) between measured concentrations and theoretical values across standard concentrations at high and normal haematocrit level.

Figure S8. Analysis of mean residual errors for caffeine (a) and paraxanthine (b) between measured concentrations and theoretical values across operators on ground and in flight onboard the aircraft.

Table S4. Dried blood spot homogeneity in weightlessness.

Figure S9. Dried blood spot collection in weightlessness and on the ground for one subject.

Figure S10A. Caffeine exposure variability according to caffeine ingested form (coffee, tea, chocolate).

Figure S10B. Paraxanthine exposure variability according to caffeine ingested form (coffee, tea, chocolate).

Figure S11A. Caffeine exposure variability according to caffeine consumption frequency (low, moderate and high).

Figure S11B. Paraxanthine exposure variability according to caffeine consumption frequency (low, moderate and high).

Figure S12A. Caffeine exposure variability on the ground for five members of our team according to the caffeine ingested form (coffee, tea, chocolate).

Figure S12B. Paraxanthine exposure variability on the ground for five members of our team according to the caffeine ingested form (coffee, tea, chocolate).

Figure S13. CYP1A2 metabolite ratio (MR) according to caffeine intake: 1: low consumption; 2: moderate consumption; 3: high consumption.

Table S5. Survey results. In this table, the exact wording of the questions is reproduced, and in parentheses, the expression used in the text to refer to each question is provided.

Data S6. Supporting information.

Derobertmasure A, Toh LS, Verstuyft C, et al. Feasibility of dried blood spot collection for caffeine pharmacokinetic studies in microgravity: Insights from parabolic flight campaigns. Br J Clin Pharmacol. 2026;92(1):35‐47. doi: 10.1111/bcp.16320

DATA AVAILABILITY STATEMENT

Data is available on request.