An Affordable Citalopram Test (ACT) for at-Home Monitoring of the Antidepressant Dosage

Introduction:: Millions of people around the world suffer from mental health disorders - a statistic that is under-reported due to the associated social stigma. It is estimated that more than 17 percent of the world population experience at least one period of major depression during their lifetime, making it one of the leading mental health disorders worldwide. Currently the standard treatment of depression is through prescribing antidepressants, which has successfully demonstrated behavioral changes in rodent models. However, the therapeutic mechanism and safety profiles has not been fully established, thus it is advised for healthcare practitioner to closely monitor their patients for unwanted side effects.

Citalopram (CTLP) is the most-prescribed antidepressant in North America and Europe due to its ability to treatment various mental disorders such as depression or panic disorder. Yet in the recent years, studies have also pointed out the danger of citalopram overdosing. In some reports researchers have shown overdosing patients may develop serotonin syndrome, a drug-induced buildup of high levels of serotonin.

As a results, studies have pointed out the importance of therapeutic monitoring in citalopram treatment for personalizing each patient’s drug dosage. At this moment, citalopram is measured through high-performance liquid chromatography, which is expensive and has a long turnaround time of up to three weeks. This deems the test unsuitable for therapeutic monitoring purposes. In this work, we report a yarn-based electrochemical sensor that is low-cost and has a short response time of less than ten seconds to address the current limitation.

Materials and Methods:: Yarn-based sensor is fabricated through the following process. First, 250 mg of PVC (Sigma Aldrich) and 500 mg of o-NPOE (Sigma Aldrich) was stirred in 2.5 mL of THF (Macron Fine Chemicals) for 3 hours. When done, 250 mg of Carbon Black VCX-72 was added to the polymer/plasticizer mixture. The ink homogeneity was achieved by mixing in pestle and mortar. The resulting carbon black ink was then dip-coated on yarn. The coated yarn was left to dry for one hour. ISM was synthesized by mixing 330 mg of PVC, 660 mg of o-NPOE, 10 mg of Sodium tetrakis(4-chlorophenyl) borate (Sigma Aldrich) and 5 mg of citalopram hydrobromide (VWR) in 3 mL of THF. The ISM was stirred overnight to ensure the mixture reaches homogeneity. Approximately 1 cm of one end of the coated yarn was dipped into the ISM mixture for 10 seconds and allowed to dry for 10 seconds afterwards, and the process was repeated two times to ensure fully coated ISM. The electrodes were then allowed to dry overnight. At last, a 3 cm heat shrink tube was placed on the electrodes, leaving 1 cm of the membrane and 1 cm of uncoated electrode exposed on either end. Heat was applied to the shrink tube by a heat gun for five to ten seconds to allow shrinking to occur while ensuring the electrode and membrane structure is not damaged by the high temperature.

Results, Conclusions, and Discussions:: Sensor calibration curve in citalopram hydrobromide solution was first established. A limit of detection (LoD) of 200.2 nM with slope of 52.3 mV/decade in the linear range of 1 mM to 0.45 μM was observed (Figure 1a and 1b). The results indicate excellent performance with sub-micromolar detection limit, with near-Nernstian behavior of the linear range fully covering both the adverse toxicity range and therapeutic range found in human urine.

The yarn-based sensor’s selectivity against interfering ions were explored next. Based on the composition of urine, we have identified sodium, potassium, calcium, magnesium, and ammonium ions as the five main interferents. Based on the results we can conclude that the ISEs were minimally affected by the interfering ions, as the ISEs were at least three orders of magnitude more selective towards citalopram over the interfering ions (Figure 1c).

Sensor performance was then assessed in spiked artificial urine solution to examine its performance in complex matrices. The ISE performance in artificial urine samples has an LoD of 572.9 nM with near-Nernstian slope of 50.1 mV/decade in the linear range of 1 mM to 1.3 μM (Figure 1d). Compared to the calibration curve obtained from citalopram hydrobromide solution, we see very small change in terms of LoD, linear range, and calibration slope, as the ISEs is in some degree affected by the abundance of interfering ions in artificial urine. However, we can also observe that the linear range still covers the toxic and therapeutic range with the LoD still in the nanomolar range. Therefore, even though interference by other non-target ions could be observed to some degree, the results indicate that the ISEs is still capable of detecting the drug concentration in complex matrices.


Acknowledgements (Optional): :

References (Optional): :