Volumetric Absorptive Microsampling (VAMS) for Therapeutic Drug Monitoring of Antiseizure Medications (ASMs) in Pediatric Patients

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for the opportunity to review this manuscript. This is a carefully executed and clinically relevant study. The development and validation of two LC-MS/MS methods for VAMS-based therapeutic drug monitoring of antiseizure medications in a paediatric population addresses a real need, and the inclusion of clobazam and its active metabolite N-desmethylclobazam is a worthwhile contribution that, as you note, has not been reported in this setting before. The real-world fingerprick sampling, the breadth of analytes covered in a single run, and the attention to ICH M10 validation parameters are all strengths. My main reservation is that several of the reported results are difficult to verify in their current form because of internal inconsistencies. The Passing-Bablok equations differ between Figure 4 and Table 7 for several analytes, the sample size for CBZ-Epoxi differs between Tables 7 and 8 (n=26 versus n=48), and the set of analytes described as requiring plasma conversion is grouped differently across the abstract, results and discussion. Since the interchangeability and conversion-factor conclusions rest directly on these figures, reconciling them is the most important task in revision. I would also encourage you to state explicitly the rule by which an analyte is assigned to the conversion pathway and to apply it consistently throughout, as TPR shows the same constant and proportional bias as CBZ-Diol and CBZ-Epoxi yet is treated as acceptable. Beyond this, the conclusions about hematocrit independence are not supported by data from this cohort, the conversion factor is derived and tested on the same dataset, and some of the stability statements read more favourably than the analyte-specific data at low QC levels would justify. These points do not undermine the value of the work, but addressing them, together with the more minor corrections noted below, would substantially strengthen the manuscript and its clinical message.

Comments for author File: Comments.docx

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript entitled “Volumetric Absorptive Microsampling (VAMS) for Therapeutic Drug Monitoring of Antiseizure Medications (ASMs) in Pediatric Patients” evaluates the applicability of VAMS as an alternative sampling strategy for therapeutic drug monitoring (TDM) of selected antiseizure medications and metabolites in pediatric patients. The study compares drug concentrations measured in capillary blood collected by VAMS with conventional plasma concentrations obtained from venous blood, using a UHPLC-MS/MS analytical platform.

The topic is clinically relevant and timely. Minimally invasive microsampling approaches are particularly attractive in pediatric populations, where repeated venous blood collection can be challenging. The manuscript addresses an important translational issue regarding the interchangeability of VAMS and plasma concentrations and provides evidence supporting the use of VAMS for several commonly monitored ASMs.

The study appears promising; however, several methodological and clinical aspects require clarification and further discussion before the manuscript can be considered for publication in pharmaceuticals journal.

Comment 1:- Table 1 need to be revised. The manuscript should provide a more detailed description of the study population, including: age distribution, sex distribution, number of patients receiving each ASM, dosage regimens, concomitant medications, and relevant clinical characteristics. Since pediatric pharmacokinetics vary substantially across age groups, stratification by age may be important. 

Comment 2:- The authors should justify whether the sample size is sufficient to support the conclusions regarding interchangeability and predictive performance. 

Comment 3:- Confidence intervals should accompany correlation and agreement statistics. 

Comment 4:- While Spearman correlation is useful, correlation alone does not demonstrate agreement. 

Comment 5:- The manuscript appropriately includes Bland–Altman and Deming regression analyses; however: acceptance criteria for agreement should be clearly defined, and the clinical relevance of observed biases should be discussed. Strong correlations may still coexist with clinically unacceptable disagreement.  

Comment 6:- The use of blood-to-plasma conversion factors for CBZ-Diol, CBZ-Epoxide, and clobazam is a key finding. However, several questions remain unanswered. For example,  
1- Are these conversion factors patient-specific or universally applicable?
2- How stable are these ratios across different hematocrit values and age groups?
3- Were external validation samples used to test the robustness of the conversion model?
Additional discussion is needed regarding the clinical implementation of these correction factors.

Comment 7:- One of the recognized advantages of VAMS is its reduced sensitivity to hematocrit effects compared with dried blood spots. Nevertheless, hematocrit can still influence analyte recovery and quantification. The manuscript should explicitly evaluate hematocrit variability among patients, potential effects on assay performance, and whether correction strategies are necessary. 

Comment 8:- The authors should state that the methods were validated according to ICH M10 guidelines. 

Comment 9:- The manuscript should discuss how VAMS-derived concentrations would affect real-world TDM decisions. 

Comment 10:- The authors should discuss reproducibility across laboratories, applicability to home sampling, and potential challenges in multicenter implementation. 

Comment 11:- The manuscript would benefit from a schematic workflow illustrating sample collection, analysis, and data comparison.

 Comment 12:-  I recommend replacing Figure 1 with a well-structured table that summarizes all key data and findings in a concise format. A tabulated presentation would improve readability, facilitate comparison among the reported parameters, and allow readers to quickly extract the most relevant information.

 

Author Response

Comment 1:- Table 1 need to be revised. The manuscript should provide a more detailed description of the study population, including: age distribution, sex distribution, number of patients receiving each ASM, dosage regimens, concomitant medications, and relevant clinical characteristics. Since pediatric pharmacokinetics vary substantially across age groups, stratification by age may be important.

We’d like to thank Reviewer# 2 for this precious suggestion. We have included in Table 1 (now Supplementary Table 5 in the revised manuscript) the clinical diagnosis and the dosing regimens for each antiseizure medication. Other demographic information including age, sex, number of patients receiving each ASM and concomitant medications were already reported.

Comment 2:- The authors should justify whether the sample size is sufficient to support the conclusions regarding interchangeability and predictive performance.

Many thanks to Reviewer#2 for raising this point. We are aware that a limitation of our study could be represented by the low number of paired samples collected for some analyte (i.e TPR and LCS). However, it’s worth to say that in our hands numbers of analysed samples were in line with previous studies in which VAMS (collected through patients’ fingerprick or by dipping tips into whole blood tubes) have been clinically tested for TDM of ASMs in a “real-world” setting:

• D'Urso A, Rudge J, Patsalos PN and de Grazia U Volumetric Absorptive Microsampling: A New Sampling Tool for Therapeutic Drug Monitoring of Antiepileptic Drugs Ther Drug Monit. 2019;41:681-692.
• Velghe S and Stove CP Volumetric absorptive microsampling as an alternative tool for therapeutic drug monitoring of first-generation anti-epileptic drugs Anal Bioanal Chem. 2018;410:2331-2341.
• Cobo-Golpe M, Paniagua-Gonzalez L, Lendoiro E, Blanco-Ces M, Lopez-Rabunal A, Abella J, Castro D, Melcon C, Fuentes P, Carballeira I, Garcia C, Gomez C, Lamas ML, Cruz A and de-Castro-Rios A Development and application of an LC-MS/MS method for 8 antiepileptic drugs and 2 metabolites using microsampling techniques (DBS and VAMS) J Anal Toxicol. 2025).
• Cancellerini C, Belotti LMB, Mohamed S, Solda M, Esposito E, Bisulli F, Mostacci B, Vignatelli L, Tinuper P, Contin M, Licchetta L and Tele-Epic study group. Fingerprick volumetric absorptive microsampling for therapeutic drug monitoring of antiseizure medications: Reliability and real-life feasibility in epilepsy patients J Pharm Biomed Anal. 2024;242:116065.

However, this aspect has been included as limitation in the Discussion section.

Comment 3:- Confidence intervals should accompany correlation and agreement statistics.

Thanks to Reviewer#2 for highlighting this point. 95% confidence intervals have been already included in Tables 7, 8 and Supplementary Table 4.

Comment 4:- While Spearman correlation is useful, correlation alone does not demonstrate agreement.

We totally agree with Reviewer #2. For this reason, the agreement between results obtained through different sampling strategies have been compared by using Bland–Altman and Deming regression analyses.

Comment 5:- The manuscript appropriately includes Bland–Altman and Deming regression analyses; however: acceptance criteria for agreement should be clearly defined, and the clinical relevance of observed biases should be discussed. Strong correlations may still coexist with clinically unacceptable disagreement. 

Many thanks to Reviewer# for raising this interesting point. For Bland-Altman plots, the acceptance criteria were that the mean difference between the results should be < 10% with no individual value > 20% (P.J. Jannetto, Chapter 8 - therapeutic drug monitoring using mass spectrometry, Mass Spectrom. Clin. Lab. (2017) 165–179). A sentence on the acceptance criteria has been added into Materials & Methods section (paragraph 4.8). Commonly, about 70-80% of the samples must lie within 20% difference. In our hands, for CBZ, CBZ-Diol, CBZ- Epoxi and CLB, 46%, 54%, 54% and 40 %, respectively, of paired samples were within the acceptance criteria. Based on this poor agreement between VAMS and plasma concentrations, we proposed to estimate plasma concentrations (ECpl) based on blood-to-plasma ratio. The obtained results showed that conversion based on blood-to-plasma ratio led to a substantial agreement between estimated versus observed plasma concentrations (Table 8). Conversely, the Passing-Bablok regression analysis still confirmed the presence of a constant and proportional bias for CBZ-Diol, -Epoxi and TPR. However, in accordance to the concept of “clinical” limits of agreement provided by Giavarina D. on the Bland Altman analysis (Giavarina D. Understanding Bland Altman analysis. Biochem Med (Zagreb). 2015 Jun 5;25(2):141-51, for all the assessed analytes in this study (including CBZ-Epoxi and TPR), the estimated mean plasma concentrations were within the therapeutic ranges adopted in our Hospital for TDM of antiseizure medications and approved by our neurologists. This point has been discussed in the Discussion section.

Comment 6:- The use of blood-to-plasma conversion factors for CBZ-Diol, CBZ-Epoxide, and clobazam is a key finding. However, several questions remain unanswered. For example, 

1- Are these conversion factors patient-specific or universally applicable?

1- Many thanks for this question. In our study, blood-to-plasma ratio (R) values were calculated as previously described and were comparable to those reported and validated in literature:

-D'Urso, A.; Rudge, J.; Patsalos, P.N.; de Grazia, U. Volumetric Absorptive Microsampling: A New Sampling Tool for Therapeutic Drug Monitoring of Antiepileptic Drugs. Ther Drug Monit 2019, 41, 681-692, doi:10.1097/FTD.0000000000000652.;

-Cancellerini, C.; Belotti, L.M.B.; Mohamed, S.; Solda, M.; Esposito, E.; Bisulli, F.; Mostacci, B.; Vignatelli, L.; Tinuper, P.; Contin, M.; et al. Fingerprick volumetric absorptive microsampling for therapeutic drug monitoring of antiseizure medications: Reliability and real-life feasibility in epilepsy patients. J Pharm Biomed Anal 2024, 242, 116065, doi:10.1016/j.jpba.2024.116065.

These conversion factors should be universally applicable; however, it depends on the tested drugs and on the microsampling devices adopted. Therefore, a comparison with current literature would be strongly recommended. Here we have used the mean of the blood-to-plasma ratios calculated for each patient and was drug specific. 

 2- How stable are these ratios across different hematocrit values and age groups?

2- This is an interesting question. However, to date this aspect has been partially evaluated in the study conducted by D'Urso and colleagues in 2019 (doi:10.1097/FTD.0000000000000652). In our study, we have observed that different blood to plasma ratio correspond to similar mean HCT % (now in Supplementary Table 3) as previously reported (D’Urso et al., 2019). This could be explained since the R values are depending on the drug type and how this is distributed between red blood cells and plasma meanwhile the HCT is patient-specific. Unfortunately, we are not able to assess whether these ratios are age-dependent.

3- Were external validation samples used to test the robustness of the conversion model?

Unfortunately, one limitation of this study could be represented by the absence of paired VAMS and plasma samples (distinct from those collected within the study), to be used as external and independent data set for further validating the adopted conversion factor. This point has been accounted in the Discussion section.

Additional discussion is needed regarding the clinical implementation of these correction factors.

As reported in the Discussion, plasma represents the standard matrix for TDM routine analysis, and discrepancies between capillary blood and plasma concentration could compromise clinician’s interpretation of VAMS results. Moreover, therapeutic ranges are often established through plasma-based assays. Therefore, converting capillary whole blood results to plasma concentrations could improve the validity of microsampling procedures and facilitate the interpretation of TDM results.

Comment 7:- One of the recognized advantages of VAMS is its reduced sensitivity to hematocrit effects compared with dried blood spots. Nevertheless, hematocrit can still influence analyte recovery and quantification. The manuscript should explicitly evaluate hematocrit variability among patients, potential effects on assay performance, and whether correction strategies are necessary.

In order to address this interesting comment, we have reported the mean HCT% of patients divided by drugs’ treatment. Thereafter, we have used the HCT% value of each patient to convert VAMS results into estimated plasma concentrations according to Equation 4 reported by Boffel L. and colleagues (Ther Drug Monit 2025, doi:10.1097/FTD.0000000000001393). Compared to the blood-to-plasma ratio conversion factor, this approach showed a worst performance for all the analytes, and led to a poor agreement between estimated versus observed plasma concentrations as assessed by Bland-Altman comparison plots (Supplementary Table 4). These results have been included now in the revised version of our manuscript.

Comment 8:- The authors should state that the methods were validated according to ICH M10 guidelines.

The ICH M10 guidelines on bioanalytical method validation and study sample analysis has been mentioned throughout the whole manuscript. Reference has been also listed (25 July 2022 EMA/CHMP/ICH/172948/2019, Committee for Medicinal Products for Human Use. Available at: https://www.ema.europa.eu/en/ich-m10-bioanalytical-method-validation-scientific-guideline. Accessed on October 4, 2022).

Comment 9:- The manuscript should discuss how VAMS-derived concentrations would affect real-world TDM decisions.

Despite their utility and potential application in different clinical settings, the main limitation of microsampling device based on capillary blood collection is the lack of specific therapeutic ranges that could compromise clinician’s interpretation of VAMS results. Moreover, therapeutic ranges are often established through plasma-based assays. Therefore, converting capillary whole blood results to plasma concentrations could improve the validity of microsampling procedures and facilitate the interpretation of TDM results. In alternative, a consensus guideline based on a systematic literature review could suggest therapeutic ranges based on VAMS-derived concentrations to be used in TDM decision. In meantime, converting capillary whole blood results to plasma concentrations could improve the validity of microsampling procedures and facilitate the interpretation of TDM results. This point has been addressed in the Discussion section.   

Comment 10:- The authors should discuss reproducibility across laboratories, applicability to home sampling, and potential challenges in multicenter implementation.

Home self-sampling represents a valid strategy to promote remote TDM and should be used as a target for the validation of VAMS in different clinical settings and for several drug classes. Moreover, home sampling represents a favourable opportunity for patients to avoid traveling toward hospitals or reference laboratories exclusively to monitor drug levels. However, it is worth noting that to avoid pre-analytical mistakes that could affect drug quantification, appropriate training on VAMS collection, storage, and shipment should be reserved for patients’ caregivers. In fact, proper education on the correct handling of these devices as well as good compliance from patients and their caregivers are essential to ensure optimal sampling and to guarantee data reproducibility across laboratories. Finally, in order to improve the use of these disposables, National Healthcare Systems should implement evidence-based policy interventions to facilitate the adoption and ensure adequate reimbursement of microsampling devices within routine clinical practice. Discussion and conclusion have been modified accordingly.

Comment 11:- The manuscript would benefit from a schematic workflow illustrating sample collection, analysis, and data comparison.

We thank Reviewer #2 for this precious suggestion. A graphical abstract illustrating the schematic workflow for sample collection, analysis, and data comparison has been realized and will be submitted alongside with the revised manuscript.

Comment 12:-  I recommend replacing Figure 1 with a well-structured table that summarizes all key data and findings in a concise format. A tabulated presentation would improve readability, facilitate comparison among the reported parameters, and allow readers to quickly extract the most relevant information.

Thanks to Reviewer #2 for this recommendation. In the revised version of our manuscript we have now introduced Table 1 that summarizes hallmarks, key data and main findings of our comparative analytical study. As consequence, Figure 1 has been moved to Supplementary material.  

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for carefully addressing the comments raised during the previous round of review. The revised manuscript is substantially improved and presents a well-conducted analytical validation of volumetric absorptive microsampling (VAMS) for therapeutic drug monitoring of selected antiseizure medications in paediatric patients. The additional methodological details, expanded discussion of stability, clearer justification of the conversion approach, and more balanced discussion of the study limitations have strengthened the manuscript considerably.

I have only a few minor comments before publication; most of these can be done during proofs. There are no additional scientific comments or issues.

The manuscript would benefit from one final round of careful English language editing to correct occasional grammatical errors, awkward sentence construction, and typographical mistakes that remain throughout the text.

Please ensure consistent terminology is used throughout the manuscript. Terms such as "antiepileptic drugs (AEDs)", "antiepileptics (AEs)", and "antiseizure medications (ASMs)" are used interchangeably. Using a single term consistently would improve clarity.

Please perform a final editorial review for minor typographical and formatting issues. A few remain in the revised manuscript (for example, "six four AEs", "underlay" instead of "underline", "worst performance" instead of "worse performance", and several spacing and formatting inconsistencies).

Consider improving the readability of several figures, particularly the chromatograms, by increasing font size and image resolution where possible.

Overall, the authors have satisfactorily addressed the previous scientific concerns. I believe the manuscript is suitable for publication following the minor corrections above.

Comments on the Quality of English Language

The manuscript is generally well written and clearly communicates the scientific content. However, a final round of careful English language editing is recommended before publication to improve grammar, sentence structure, and overall readability. Minor typographical errors, inconsistent terminology, and occasional awkward phrasing remain throughout the manuscript. 

Reviewer 2 Report

Comments and Suggestions for Authors

I see that the authors have appropriately answered all questions and concerns. The manuscript could be accepted in its current form