DBS have been used in newborn screening since the sixties [33] and in many other bioanalytical fields, such as elemental analysis [16,34], nucleic acids research [35], forensic toxicology [36], proteomics, genomics, and metabolomics [37]. DBS have also been used for the TDM of several classes of drugs, such as antiepileptics drugs (AEDs) [38,39,40], antiretrovirals [41], anticancer drugs [42,43,44], immunosuppressants [11,45], antibiotics [9], antituberculosis [46], and neuroactive drugs [47,48,49,50]. Using DBS is a very convenient option compared with traditional sampling, but drug concentrations in capillary blood may be slightly different from those measured in venous blood [51]; therefore, drug-specific reference intervals are needed to implement DPS-based TDM in clinical practice [12]. Generally, in the dry matrix, given the absence of water, drugs are more stable, but the stability in DBS must be verified for each specific analyte. For example, the stability of ceftolozane is more limited in DBS than in liquid blood [9]. Different types of filter paper can show different matrix effects on the analytes [52]. Interestingly, different matrix effects in the quantification of antipsychotic drugs and their metabolites were observed when results obtained with various cellulose-based untreated filter papers—such as the Whatman^® 903 Protein Saver Card or the Fast Transient Analysis (FTA^®) Drug Metabolism and Pharmacokinetic (DMPK) type C Card—were compared [53].

The main issue for drug quantification in DBS is the hematocrit (Hct) effect. Hct is the percentage of red blood cell volume in blood and strongly influences blood viscosity, blood droplet volume, blood droplet migration on the paper substrate, drying time, homogeneity, and spot size, affecting the accuracy and precision of drug quantification. Different Hct values may compromise the reproducibility of the analysis because of the uneven migration of cells and fluids on the paper structure [54,55]. Many attempts have been made to solve the Hct issue [56]. Abu-Rabie et al. (2015) showed that the overall bias given by Hct includes an Hct-based area bias, an Hct-based recovery bias, and an Hct-based matrix effect bias [57]. One example is whole spot analysis, which could overcome the problem of uneven blood distribution on the paper substrate, but the exact volume of blood drawn must be known [58].

The quantification of drugs in microsamples can also be strongly influenced by pre-analytical variables, such as the type of solvent used for drug extraction [42,44,45,48,59] or the way the internal standard (IS) is added to the samples [57,60]. Despite some novel techniques to address the IS having been proposed in the literature, such as the TouchSpray^® [60] or the post-column infusion modality (PCI-IS) [61,62], the IS is usually pre-diluted into the extraction solvents. Placing the IS on the DBS and drying it before extraction should be considered the most reliable method to verify drug recovery. Sonication, heating, and the addition of water before extraction for partial rehydration of samples can greatly improve extraction efficiency. Interestingly, it has been observed in several cases that paper substrates can retain proteins, lipids, and phospholipids, providing very clean extracts from dried microsamples, thus improving the performance of LC-MS/MS in analyzing the corresponding liquid microsamples [63].

Several methods for TDM of different classes of drugs are reported in the literature. For example, TDM is used for different AEDs to optimize dosing in individual patients. DBS dosing appears to be a viable alternative to conventional TDM on plasma. Pohanka et al. (2014) [40] developed and validated an LC-MS method for the measurement of valproic acid in dried blood spots. The use of blood samples ranging in size from 20 to 100 μL did not yield significantly different valproic acid concentrations, and the method proved robust in the 30–60% hematocrit range. A comparison between DBS and plasma was performed, and plasma concentrations were significantly higher than DBS, emphasizing the need to create method-specific reference ranges for each analysis. LC-MS/MS methods for the quantitation in DBS were developed also for topiramate [64], phenobarbital [65,66], lamotrigine [66,67], rufinamide [68], clobazam [69], clonazepam [69], levetiracetam [66], and carbamazepine [66,70]. DBS-based methods have also been applied in the TDM of anticancer drugs. Recently, Poetto et al. (2021) [71] developed and validated a dried blood spot LC-MS/MS method for the TDM of palbociclib, ribociclib, and letrozole in patients affected by cancer, and they observed a positive correlation between DBS and plasma concentrations for the three drugs. Berm et al. (2015) [48] presented a method for therapeutic drug monitoring of the tricyclic antidepressants amitriptyline, nortriptyline, imipramine, clomipramine, and their active metabolites in DBS using LC-MS/MS. The authors observed that a low hematocrit (≤30%) was associated with a negative bias (≥15%) for all analytes. In contrast, punching the blood spot sample from the perimeter instead of the center was associated with a positive bias. A good correlation was found between the patients’ plasma and DBS samples for all analytes except clomipramine.

DBS devices have been widely used in the TDM of immunosuppressants. Veenhof et al. (2023) [72] conducted a pilot proficiency test for the microsampling of immunosuppressants (tacrolimus, cyclosporine, everolimus, sirolimus, mycophenolic acid) involving 14 laboratories from seven countries in three rounds of proficiency testing. Immunosuppressant microsampling methods showed high interlaboratory variation compared with the whole blood methods, underscoring the need for harmonization and standardization. Proficiency testing should be routinely performed for laboratories using immunosuppressant microsampling techniques in patient care. In fact, Veenhof et al. (2019) [73] applied a DBS assay to measure sirolimus and everolimus in transplant patients. Passing–Bablok regression showed no significant differences between whole blood and DBS, but the limits or clinical significance were not reached (77.3% and 61.5%, respectively). In an effort to reduce the hematocrit effect and volume problem that plague DBS, devices capable of collecting definite volume samples have been designed [74], such as disposable low-cost viable capillaries [75] and DBS with metering capillary channels. Alternatives for an Hct-independent determination of drugs in blood microsamples are the Volumetric Absorptive Paper Minidiscs (VAPD-mini) [76] and the Hemapen^® [77].

A popular technique to control the volume of blood microsamples is volumetric absorptive microsampling (VAMS), a device consisting of a globular hydrophilic tip mounted on a plastic tip to collect a fixed volume of sample [78,79,80]. In 2014, Neoteryx commercialized a microdevice called Mitra^®, based on the principle of VAMS [80], which has been designed to present all the advantages of the DBS technique without the effect of hematocrit, simplifying the workflow for the analysis of whole blood samples [81]. Different configurations of VAMS device are available, allowing 10, 20, and 30 µL of whole blood to be collected [80]. A finger or heel prick is made, then the adsorbent sampling tip is placed in contact with only the surface of the head of the tip in the blood drop [80,82]. The tip is inserted into the blood drop, allowing adsorption by capillarity. The tip must be held in contact with the blood drop for approximately 2–3 s to allow complete filling [82]. Contact times longer than 6 s may alter the volume collected by overfilling the tip. The VAMS method results were accurate and reproducible, even under home sampling conditions [83]. From the standpoint of home sampling, where collection is carried out without the help of trained health care personnel, adequate training on how to sample with VAMS is extremely critical to ensure a good sample quality. An often-adopted solution, which has proven successful, is to provide training video tutorials and instructions online [80,82,84]. Several studies have demonstrated the low impact of Hct on the analytical performance of VAMS [85,86,87,88,89,90].

Again, a comparison of plasma/blood and VAMS methods is needed to apply them in the clinical setting. For cannabidiol (CBD) and its main metabolites, it has been shown that concentrations in VAMS devices and plasma are not significantly different [91,92]. Therapeutic drug monitoring of blood levels of cannabinoids is crucial for optimizing the medical cannabis therapy, and the use of microsampling devices could facilitate the widespread adoption of this clinical practice, as well as simplify the sampling in patients who are not compliant with venipuncture. Another technique for collecting blood microsamples is three-dimensional (3D) dried blood spheroids (3D-DBS), a device based on hydrophobic papers, in contrast to traditional planar (2D) hydrophilic cellulose-based papers. Cellulose is functionalized with trichloro(3,3,3-trifluoropropyl)silane. Aqueous blood samples are deposited on the surface as droplets, leading to the formation of 3D-DBS. The blood spheroids form a barrier between the analytes and air that protects the analytes from oxidative degradation and thermal conduction [93]. Paper functionalization has recently been exploited also for the production of molecularly imprinted-interpenetrating polymer network (MI-IPN) devices [94].