Fingertip blood microsampling with hematocrit-adjusted conversion shows promise for transplant drug monitoring

For the hundreds of thousands of people living with transplanted organs, the daily routine of immunosuppressive drugs is a tightrope walk. Too little tacrolimus or mycophenolic acid in the blood, and the immune system may begin attacking the new organ. Too much, and the same drugs that protect the graft can poison the kidneys, suppress the bone marrow, or invite life-threatening infections. Keeping patients on that tightrope has long required repeated trips to the clinic, needles in the arm, and vials of blood. Now a team at National Taiwan University Hospital has shown that a tiny fingertip-sized microsample, analyzed with the right mathematical corrections, can stand in for conventional venous blood draws in monitoring two of the most important transplant drugs.

The study, published in the Journal of Pharmacy & Pharmaceutical Sciences, focused on volumetric absorptive microsampling, or VAMS, a technique in which a small absorbent tip wicks up a fixed volume of blood from a finger prick. The device looks almost trivially simple, but the chemistry behind it is not. Once blood dries on the tip, the concentration of a drug measured from the whole sample may not match what clinicians would see in plasma, the matrix on which most therapeutic targets for mycophenolic acid are based. The reason lies in how the drug partitions between plasma and red blood cells, and that partitioning is governed largely by hematocrit, the fraction of blood occupied by red cells.

Tacrolimus and mycophenolic acid sit at opposite ends of this problem. Tacrolimus is a calcineurin inhibitor that suppresses T-cell activation, and it is extensively taken up into erythrocytes, which is why whole blood has always been the recommended monitoring matrix. Mycophenolic acid, the active metabolite of mycophenolate mofetil and mycophenolate sodium, behaves differently. It inhibits an enzyme called inosine monophosphate dehydrogenase, shutting down lymphocyte proliferation, but the molecule itself stays almost entirely in plasma, where roughly 97 percent of it is bound to proteins. Measuring it from whole blood therefore gives a number that is not directly comparable to the plasma concentrations on which decades of clinical experience rest.

To bridge that gap, the researchers enrolled adult transplant recipients, mostly kidney and lung transplant patients, who were already receiving mycophenolate therapy, and collected paired samples of plasma, whole blood, and VAMS microsamples between June and October 2023. Twenty-one paired samples were analyzed with a validated liquid chromatography-tandem mass spectrometry method capable of quantifying both drugs simultaneously from a single microsample. The analytical performance was strong. Calibration curves were linear across ranges of 10 to 20,000 nanograms per milliliter for mycophenolic acid and 0.5 to 500 nanograms per milliliter for tacrolimus, with correlation coefficients exceeding 0.99, and accuracy fell within the accepted 85 to 115 percent window at nearly all quality-control levels.

With the assay validated, the team turned to the harder question: how to convert a VAMS whole-blood concentration into a plasma-equivalent value for mycophenolic acid. They compared five strategies, including two variations of a formula derived from dried blood spot work, one VAMS-specific linear model, a multiple regression model, and a physiologically based population model. The simplest approach, which assumed a fixed hematocrit of 0.45 liters per liter for every patient, performed poorly, achieving agreement within 20 percent in only 19 percent of samples and carrying a mean bias above 30 percent. The regression model that included hematocrit as a covariate but lacked an interaction between hematocrit and concentration also showed marked proportional bias.

The clear winner was an individualized correction that divides the VAMS concentration by one minus the patient’s own measured hematocrit and then multiplies by the plasma protein binding fraction of 0.97. This formula achieved clinical agreement in 86 percent of samples, with a mean Bland-Altman bias of just over 3 percent and tight predictive error metrics. A physiologically derived population model performed nearly as well, and a VAMS-specific model with a concentration-hematocrit interaction term actually produced slightly better numerical statistics, but the researchers judged the individualized formula the best balance of accuracy, interpretability, and ease of calculation for routine clinical use. Deviations from plasma values clustered mainly at very low concentrations, below roughly 488 nanograms per milliliter, a range that lies beneath the therapeutic exposure window and rarely changes clinical decisions.

Tacrolimus told a different story. Because the drug distributes so heavily into red cells, VAMS concentrations correlated strongly with conventional whole-blood values without any hematocrit correction at all. Passing-Bablok regression showed a slope close to unity, and the modest negative bias of about 18 percent remained within predefined acceptance limits. In practical terms, that means a single finger-prick microsample could replace a venous draw for routine tacrolimus trough monitoring, while the same sample simultaneously delivers a mycophenolic acid measurement once the individualized hematocrit adjustment is applied.

The clinical payoff became vivid in two lung transplant recipients who developed leukopenia, a dangerous drop in white blood cells, while taking the drug combination. Using eight-point VAMS sampling across a dosing interval, the team reconstructed full area-under-the-curve profiles for mycophenolic acid. The complete profiles yielded AUCs of 28.9 and 23.1 milligrams per hour per liter, near or below the lower therapeutic threshold of 30, while conventional three-point schemes underestimated exposure by 16 to 35 percent. Notably, the dense sampling captured secondary peaks at six to eight hours after dosing, the signature of enterohepatic recirculation, which sparse sampling misses entirely. Despite these subtherapeutic AUCs, both patients exhibited leukopenia, most likely due to the combined myelotoxic effects of concomitant immunosuppressants and anti-infective agents rather than excessive mycophenolic acid exposure, a distinction that full-profile monitoring can help clinicians make.

The authors are careful about the limits of their work. The cohort was small, the patients had hematocrit values well below the general population average, and all microsampling was performed by trained professionals rather than by patients themselves. VAMS specimens were also generated from venous whole blood rather than true finger-stick capillary samples, although published data suggest negligible capillary-venous bias for these drugs. The eight-point sampling scheme, while it establishes a high-resolution pharmacokinetic reference, is unlikely to be practical for everyday monitoring; its real value is as a foundation for developing limited sampling strategies and population pharmacokinetic models that could predict full exposure from far fewer time points.

Even with those caveats, the study marks a meaningful step toward decentralizing transplant drug monitoring. A validated microsampling workflow that handles both tacrolimus and mycophenolic acid in one sample, paired with an individualized hematocrit correction that any clinical laboratory can compute, could eventually allow patients to monitor their immunosuppression from home, mailing dried microsamples to the lab. The researchers suggest that future work should evaluate home-based self-sampling, validate simplified sampling strategies derived from VAMS data, and test the approach across broader transplant populations. If those steps succeed, the days of the transplant tightrope walk may become considerably less precarious.

Subject of Research: Hematocrit-adjusted volumetric absorptive microsampling for tacrolimus and mycophenolic acid therapeutic drug monitoring in transplant recipients

Article Title: Evaluation of hematocrit-adjusted conversion strategies for mycophenolic acid and tacrolimus monitoring using volumetric absorptive microsampling in lung and renal transplant recipients

Article References: Juan, L., Cheng, C.-N., Chang, W.-C., Chiu, H.-H., Lu, C.-W., Hsu, H.-H., Kuo, C.-H., & Tseng, Y.-J. (2026). Evaluation of hematocrit-adjusted conversion strategies for mycophenolic acid and tacrolimus monitoring using volumetric absorptive microsampling in lung and renal transplant recipients. Journal of Pharmacy & Pharmaceutical Sciences, 29, Article 16123. https://doi.org/10.3389/jpps.2026.16123

Image Credits: AI Generated

DOI: 10.3389/jpps.2026.16123

Keywords: volumetric absorptive microsampling, mycophenolic acid, tacrolimus, therapeutic drug monitoring, hematocrit, organ transplantation, LC-MS/MS, pharmacokinetics, area under the curve, immunosuppression, enterohepatic recirculation, finger-prick sampling