Therapeutic Drug Monitoring of Voriconazole in AIDS Patients

Tingting Chen Quanzhou First Hospital A liated to Fujian Medical University Zhiqiang Lin Quanzhou First Hospital A liated to Fujian Medical University Huatang Zhang Quanzhou First Hospital A liated to Fujian Medical University Qingquan Zhang Quanzhou First Hospital A liated to Fujian Medical University Limian Hong Quanzhou First Hospital A liated to Fujian Medical University Shuifa Wu (  shfw0116@126.com ) Quanzhou First Hospital A liated to Fujian Medical University https://orcid.org/0000-0002-8023-1874


Background
The safety and e cacy of Voriconazole in Acquired Immune De ciency Syndrome (AIDS) patients is di cult to guarantee. In this study, Therapeutic Drug Monitoring (TDM) of Voriconazole in AIDS patients was investigated with the aim to further verify the signi cance of voriconazole TDM in AIDS patients and to explore more strategies to improve individualized medication.

Methods
The data of AIDS patients who underwent voriconazole TDM in our hospital from May 2018 to August 2021 were collected. The basic information of patients, the results of voriconazole TDM, the individualized intervention, the affecting factors of voriconazole concentration were analyzed, as well as the relationship between voriconazole trough concentration and safety.

Results
A total of 46 tests of voriconazole TDM were performed in 28 AIDS patients. Only 57.14% patients reached the therapeutic range at rst TDM, and 87.50% patients reached the therapeutic range after intervention based on rst TDM. 21.43% patients develop voriconazole-related Adverse Drug Reactions (ADRs), and ADRs were mostly occurred when voriconazole concentration is above 5.0 µg/mL. Spearman correlation coe cient r s was calculated to be 0.729 for voriconazole trough concentration and the incidence of ADRs, exhibiting a signi cant, positive linear correlation (P=0. 017). 50% patients had polypharmacy and drug interactions are common. For example, rifampicin can signi cantly reduce the plasma concentration of voriconazole. Multiple linear regression analysis showed Hypoproteinemia was a signi cant factor affecting voriconazole trough concentration(P=0.006).

Conclusion
AIDS patients usually have a low attainment rate of voriconazole trough concentration after initiation of standard dosing regimen. The affecting factors seem multifactorial and complex, of which hypoproteinemia is of great signi cance. Meanwhile, we need to be alert to the effects of drug interactions. The incidence of voriconazole related ADRs is high, mostly occurring when voriconazole concentration is above 5.0 µg/mL. Therefore, TDM can provide meaningful guidance for dosage optimization of voriconazole, and the dosage adjustment method in Chinese Guideline is applicable for the population of AIDS patients.

Background
Voriconazole, a triazole antifungal agent, has been widely used in clinical practice. Voriconazole exhibits nonlinear pharmacokinetics in vivo, and the therapeutic window is narrow. TDM can improve the safety and e cacy of voriconazole therapy [1,2]. Due to the special disease status and the polypharmacy in AIDS patients, the individualized dosing of voriconazole is more complicated, and this is an issue that deserves more attention and further research is critical. Unfortunately, researches in this regard are limited, particularly the voriconazole TDM in General clinical information and purpose of voriconazole administration From May 2018 to August 2021, 28 AIDS patients with TDM-based dosage adjustment of voriconazole were admitted to our hospital, 24 (85.71%) were male and 4 (14.29%) were female, mean age was 39.71±14.14 years (22-75 years), mean weight was 52.25±11.22 kg (33-75 kg), mean serum albumin (ALB ) 28.58±5.03 g·L -1 (19.10-44.50 g·L -1 ), mean glutamate transaminase (ALT) 38.68±40.96 U·L -1 (2.00-163.00 U·L -1 ), mean glutamic aminotransferase (AST) 70.29±122.20 U·L -1 (9.00-666.00 U·L -1 ), and all patients presented normal renal function. The purposes of voriconazole administration were shown in Table 1.

Voriconazole TDM results
Of the 28 AIDS patients, 26 were prescribed voriconazole at rst-time and given a loading dose of 6 mg·kg -1 q12h followed by a maintenance dose of 4 mg·kg -1 q12h intravenously. 2 patients had taken voriconazole before admission, and they continued take voriconazole orally with a maintenance dose of 4 mg·kg -1 q12h after admission. Blood samples were collected 0.5 h before the 5th dose at the earliest, and the mean sampling time was 4.25 d (2-11 d) after initiation of voriconazole therapy. A total of 46 tests of voriconazole concentration were performed in 28 patients, of which 21 patients had only 1 test of voriconazole concentration and 7 patients had multiple tests of voriconazole concentration, with a mean number of tests 3.6 (2-7). The mean trough concentration at rst voriconazole TDM from all the 28 AIDS patients was 4.17 ± 2.70 μg/mL (0.2-9.6 μg/mL). The distribution of voriconazole TDM results were shown in Table 2.
Intervention based on voriconazole TDM There were 12 patients who did not reach the target range at their rst voriconazole TDM, and dosing regimens were intervened according to the therapeutic effect and TDM results. Five patients reached the standard in the second TDM after intervention, among which three patients underwent voriconazole dose reduction, one patient underwent voriconazole dose escalation, and one patient discontinued rifampicin used in combination. No dose adjustments were required in three patients according to guideline recommendation. Voriconazole was replaced by itraconazole in two patients. One patient discontinued voriconazole due to drug fever. Another patient initially treated with voriconazole for empirical treatment was switched to amphotericin B. The results of voriconazole TDM and corresponding interventions were shown in Table 3.

Trough concentration and safety of voriconazole
Among the 28 patients, 6 patients(21.42%) encountered voriconazole related ADRs, including 3 cases of hallucinations, 1 case of visual disturbance, and 1 case of anxiety and insomnia. Of the 6 patients, there were 5 patients that the voriconazole trough concentrations were larger than 5 μg/mL, accounting for 83.33%. The mean voriconazole trough concentration in ADR group (6 cases) and non ADR group (22 cases) was (7.12±2.27) μg/mL and (3.37±2.23) μg/mL, respectively. Statistical difference could be observed between the two groups (P=0.001). Correlation between the two groups was examined using Spearman Correlation analysis. Spearman correlation coe cient r s was calculated to be 0.729 for voriconazole trough concentration and the incidence of ADRs, exhibiting a signi cant, positive linear correlation (P=0. 017). The distribution of voriconazole trough concentrations and the occurrence of ADRs were shown in Table 4.

Affecting factors on voriconazole trough concentration
Multiple linear regression using stepwise procedure were conducted to investigate the relationship between the voriconazole trough concentration and the affecting factors including gender, age, body weight, drug interactions, ALB and Child-Pugh classi cation. The nal results of multiple linear regression analyses were summarized in Table 5.
Multiple linear regression analysis showed that gender (P=0.044) and ALB (P=0.006) had the greatest impact on voriconazole trough concentration. The R 2 =0.406 indicated that 40.6% variations of voriconazole trough concentration could be explained by drug interaction, ALB, gender, weight, and Child-Pugh classi cation. Since the majority of the AIDS patients included in this study were male, the impact of gender was not referential.

Discussion
Candida infection and Cryptococcus infection are prevalent fungal opportunistic infections in AIDS patients, and in southern China, Talaromyces marneffei(T. marneffei)-infection is also common [9]. For Candida infections, echinocandins and uconazole are preferred over voriconazole [10]. For Cryptococcus neoformans infection, amphotericin B, ucytosine and uconazole are often the preferred choice [11,12]. For T. marneffei-infection, a fatal systemic mycosis, the recommended medication regimen is induction treatment with amphotericin B followed by a switch to maintenance oral itraconazole [13]. Therefore, voriconazole is not routinely recommended as the rst choice for the common opportunistic fungal infections in AIDS patients. Previous studies have indicated that most isolates of T. marneffei are susceptible to voriconazole [14,15]. Series studies evaluating the e cacy and the safety of voriconazole to treat patients with T. marneffei infection have demonstrated that voriconazole is an effective, well-tolerated therapeutic option for this kind of disease [15]. Voriconazole is also active against Candida and Cryptococcus. Therefore, voriconazole is a guideline-recommended alternative in Candida infection, Cryptococcus infection and T. marne i infection [10][11][12][13]. In this study, we collected HIV cases for more than 3 years, voriconazole was frequently used as an alternative treatment for patients who were intolerant to other antifungal drugs. Thus, the number of HIV cases was relatively small in our study. Nevertheless, the rational use of voriconazole in AIDS patients is essential and of importance. 78.57% AIDS patients used voriconazole for targeted therapy and 21.43% for empirical treatment (Table 1). AIDS patients are a special population with special disease status and polypharmacy, so the individualized medicine of voriconazole in a clinical setting become more challenging. Series studies have investigated the application of voriconazole TDM in special populations, including children [16], the elderly [17], patients with cirrhosis [11], patients with tuberculosis [18]. However, no literature has reported the application of voriconazole TDM in AIDS patients, which is the outstanding highlight of this present study. Moreover, the clinical pharmacist was involved in the whole treatment, for example, took part in selecting appropriate drugs and formulating treatment regimens for AIDS patients, conduct voriconazole TDM, guide the dose adjustment of voriconazole based on TDM and monitoring of ADRs. Therefore, clinical pharmacist can play an important role in the treatment of AIDS. Besides, clinical pharmacist has emerged as an indispensable member of the multidisciplinary diagnosis and treatment team.
The therapeutic range of trough concentration and the applicable population for voriconazole TDM are still controversial [19]. The voriconazole treatment regimens for 26 cases are all given a loading dose and a maintenance dose based on body weight. The maintenance dose was not halved in 2 patients with hepatic function Child-Pugh class B. Blood samples from the 28 cases were collected 0.5 h before the 5th dose at the earliest, and the average sampling time was 4.25 d after initiation of voriconazole therapy, the timing of blood sampling in the 28 cases all met the guideline requirements [4][5][6][7][8], which ruled out the in uence of improper sampling on the TDM results. However, the rst results of voriconazole TDM showed that the concentration attainment rate is only 57.14% (Table 2). Recently, there have been plenty of researches about the application of voriconazole TDM on patients with invasive fungal infections, but the patients in our study have a lower attainment rate of concentration compared with other research [20]. Dose adjustment based on TDM is an important element of individualized drug monitoring, and domestic and foreign clinical guidelines have different recommendations. Chinese guideline recommend that 20% maintenance dose can be reduced when the concentration reach above the upper limit and below 10.0 µg/mL and no grade 2 or higher adverse events occurred [8]. Moreover, 50% maintenance dose can be increased when the concentration are below the lower limit or when voriconazole treatment shows suboptimal e cacy [8]. In this study, AIDS patients who did not achieve target concentration range were adjusted to individualized regimens based on TDM results, including avoidance of drugs carrying a high risk of drug-drug interactions, appropriate dose increment or reduction, or switching to alternative regimens. There are three patients in this study reduced the dose by 20% (Table 3). Also, the dose was increased by 20% in another patient since a trough concentration of 0.9µg/mL was measured from the rst TDM (Table 3). After individualized intervention, the trough concentration of voriconazole in aboved four patients all reached the standard range as second TDM indicated( Table 3), demonstrating that the Chinese dose adjustment guideline is applicable to the population of AIDS patients. Furthermore, more attention should be paid to the patient's compliance and prompt withdrawal of the interacting drugs other than dose adjustment, according to British guideline [5]. After discontinuation of rifampin, the voriconazole concentration of one patient in our study also rise to the lower limit without dose adjustment (Table 3). Consequently, the attainment rate of trough concentration increased signi cantly after intervention based on rst TDM results, from 57.14-87.50% (Table 2). Thus, there is signi cant need to perform TDM for AIDS patients using voriconazole.
A meta-analysis showed that the risk of developing drug toxicity could be signi cantly increased when trough concentration of voriconazole exceed 6.0 µg/mL. 2 During hospitalization, the incidence of voriconazole associated ADRs is 21.43% in our study (Table 4). 83.33% of the patients suffering from ADRs had trough concentrations above 5.0 µg/mL, manifesting as hallucinations, visual disturbances and drug-related hepatitis( Table 4). The mean trough concentration of voriconazole in the ADR group was (7.12±2.27) µg/mL. It is worth mentioning that the incidence of ADR exhibits positive correlation with the trough concentration level of voriconazole (P=0. 017).
There are many multifactorial and complex affecting factors that can in uence the individual differences in clinical application of voriconazole, and the factors that have been con rmed include genotype, age, gastrointestinal absorptions, pathophysiological status and drug interaction, of which Cytochrome P450 2C19(CYP2C19) genetic polymorphisms and drug interaction are the two most important affecting factors [17,21].
The Cmax and AUC values of voriconazole in poor CYP2C19 metabolizers are 2-5 times higher than normal CYP2C19 metabolizers [22]. The CYP2C19 genetic polymorphism in uences the steady-state level of voriconazole and is a signi cant affecting factor contributing to the highly variable pharmacokinetics of voriconazole [23]. The detection of CYP2C19 genetic polymorphism is recommended for patients using voriconazole by Clinical Pharmacogenetics Implementation Consortium (CPIC) [24]. It is worth mentioning that the distribution characteristics of CYP2C19 genetic polymorphisms can differ by race and ethnicity. About 20-30% Asian populations are slow metabolizers, while only about 2-3% Caucasians are slow metabolizers [25]. For 32.14% of the AIDS patients in this study, the mean trough concentration of voriconazole skewed toward higher values (Table  2), and the observed phenomenon may be blamed for genotype. Nevertheless, the present study suffered from a drawback that the patients were not subjected to detection of CYP2C19 genetic polymorphisms.
Drug interaction is another important affecting factor on steady-state level of voriconazole. AIDS patients are immune-suppressed and susceptible to opportunistic infections, so polypharmacy is often required. Thus, there is clearly a need to be more alert to the potential drug interactions. Voriconazole is metabolized mainly in liver, both an inhibitor of CYP2C19, CYP2C9, CYP3A and an enzyme substrate. So voriconazole can easily interacts with other CYP enzyme substrates, inhibitors or inducers. The common drugs observed drug-drug interactions with voriconazole include antiretroviral drugs, antiepileptic drugs, PPIs, rifamycin, antibacterial drugs, glucocorticoids, calcium channel antagonists, sedative hypnotics, antiarrhythmic drugs and so on. CYP inducers can reduce systemic exposure to voriconazole and lead to treatment failure when trough concentration of voriconazole are below the therapeutic range [26]. Rifampicin, a CYP3A4 inducer, can dramatically decrease the Cmax and AUC of voriconazole in vivo [27]. In our study, a patient diagnosed co-infection with mycobacterium tuberculosis and T. marneffei was co-administrated with voriconazole during quadruple antituberculosis therapy for pulmonary tuberculosis (isoniazid, rifampin, ethambutol and pyrazinamide). The dosage of voriconazole and the timing of sampling were in accordance with guideline recommendations. Unexpectedly, the voriconazole trough concentration was measured to be 0.2µg/mL, which was far below the lower limit of the therapeutic range. After discontinuation of rifampicin for 7 days, the voriconazole trough concentration rised to 1.0µg/mL without dose adjustment, which was within therapeutic range. Our results further con rmed the potent enzyme-inducing effect of rifampicin on the trough concentration of voriconazole. In contrast, co-administration with CYP inhibitors may increase systemic exposure to voriconazole and alter its safety pro le [2]. When trough concentration of voriconazole is above the therapeutic range, an increased incidence of ADR may occur during voriconazole therapy [2]. The most common drugs in combination with voriconazole in this study are PPIs. Surprisedly, there were no effects of esomeprazole or omeprazole on voriconazole trough concentration could be signi cantly observed between the no-combination group and the co-administered with PPIs group(P=0.76). Since this study is a retrospective study in a single center, and the number of cases is relatively small, there is an inherent problem in studying drug interactions, especially co-administration of CYP inhibitors including PPIs.
The speci c pathophysiological status of AIDS patients is considered another important factor in uencing the steady-state level of voriconazol in blood. AIDS patients are prone to develop hypoproteinemia during treatment, which may also bring about alterations in pharmacokinetic and pharmacodynamic properties of voriconazole. It is well known that the bound drug-albumin complex is the storage form of a drug, and the complex can gradually release free drug when the blood concentration decreases. On one hand, only the unbound or free fraction of a drug can cross the biological membrane and exert pharmacological effects within the body. On the other hand, only the unbound or free fraction of a drug can be cleared from the body. Therefore, serum ALB level will have a signi cant effect on the apparent volume of distribution (Vd) and clearance (CL) of highly albumin-bound drugs.
Some studies have demonstrated that hypoproteinemia is an in uential factor affecting trough concentration of voriconazole [28]. On the one hand, hypoproteinemia can result in increased Vd and CL of a drug, potentially resulting in under-dosing and consequent treatment failure. On the other hand, hypoproteinemia can alter the plasma protein-binding rate of voriconazole, potentially resulting in increased concentration of unbound voriconazole and consequent ADRs, even when the trough concentration of voriconazole is within normal range [28]. One patient with a trough concentration within the therapeutic range, mainly manifested as anxiety and insomnia (Table 4), requiring consideration of hypoproteinemia(ALB=28.4g/L). The mean ALB of AIDS patients in this study was (28.58±5.03) g·L −1 , and the ALB level appeared to be a important factor affecting steady-state level of voriconazole in blood (

Conclusion
In summary, the present study indicated the di culty of predicting voriconazole concentration in blood without TDM. AIDS patients usually have low attainment rate of voriconazole concentration after initiation of standard dosing regimen, and the affecting factors seem multifactorial and complex. Hypoproteinemia is a signi cant affecting factor, as well as drug interactions. The incidence of voriconazole related ADRs in AIDS patients is high, mostly occurring when the concentration of voriconazole is above 5.0 µg/mL. Accordingly, the implementation of pharmaceutical care based on TDM is vital to enhance the e cacy and safety of voriconazole therapy, which can guide the dosage optimization of voriconazole. Last but no least, dosage adjustment method in Chinese Practice Guideline is applicable for the population of AIDS patients. However, there are some limitations in our study. First of all, the number of cases is small. Second, the detection of CYP2C19 genetic polymorphisms is not performed. Therefore, more efforts should be made to further investigate the clinical implications of voriconazole TDM in AIDS patients, and to explore more strategies to improve individualized medication.