Metabolizing profile of the cytochrome pathway CYP2D6, CYP3A4 and the ABCB 1 transporter in Spanish patients affected by Gaucher disease
Abstract
Several therapeutic options are available for type 1 Gaucher disease (GD1), including enzymatic replacement therapy (ERT) and substrate reduction therapy (SRT). Eliglustat is a selective inhibitor of glucosylceramide synthase that is extensively metabolized by CYP2D6 and, to a lesser extent by CYP3A4; it is also an inhibitor of the P-gp transporter. The aim of this study is to evaluate the metabolizer profile of these cytochrome isoforms in 61 GD1 patients, and to analyze interferences with concomitant therapies. Patients were selected from the Spanish Gaucher Disease Registry considering clinical data, GBA genotype, severity score index, comorbidities, concomitant drugs, type and response to therapy and adverse effects. The polymorphisms of CYP2D6, CYP3A4 and three ABCB1 transporter variants were analyzed by Polymerase Chain Reaction (PCR). The most frequent metabolizer profile was extensive or intermediate for CYP2D6, extensive for CYP3A4*1B and CYP3A4*22 and normal activity for ABCB1. Correlations between metabolizer profile and other variables were analyzed by multiple regression study. Twenty-eight patients received ERT, 17 eliglustat and seven miglustat. Forty-two patients (68.8%) had associated diseases and 54.5% were taking daily concomitant medication. Nine patients under eliglustat therapy received concomitant drugs that interact with the CYPs and/or ABCB1, five of these did not reach therapeutic goals and three presented mild or moderate adverse effects (headache and gastrointestinal disorders). Detailed analysis in four patients with TTT haplotype, corresponding to lack of activity of the transporter, was performed. In order to apply personalized medicine and avoid interferences and adverse effects, the individual CYP metabolizer profile and transporter must be considered when choosing the concomitant medication and/or making dose adjustments.
1. Introduction
Gaucher disease (GD) (OMIM#230800) is a recessive inherited error of lysosomal glycolipid metabolism that is caused by homozygous or biallelic pathogenic variants in the GBA1 gene (MIM:6064639), which encodes the lysosomal enzyme glucocerebrosidase (GluCer, acid β-glucosidase, EC 3.2.1.45) [1–4]. Pathogenic variants in GBA1 lead to varying degrees of deficient and/or defective GluCer function, and to the lysosomal accumulation of excess glucosylceramide, glucosyl-sphingosine and other sphingolipids in macrophages. GD has a pan-ethnic distribution in Spain, its prevalence is about 1/100,000 [5]. The Spanish Gaucher Disease Registry (SGDR) has been working since 1993 compiling demographic, clinical, genetic, analytical and imaging data of Spanish GD patients.
GD is characterized by a wide degree of clinical heterogeneity. Traditionally, GD is classified into three types, based on the presence and degree of neuronopathic affectation, however, the tendency is to consider it as a continuum of phenotypes [2]. In general, the clinical manifestations are related to the effects on the mononuclear-macrophage system and are characterized by enlargement of the liver and spleen secondary to accumulation of macrophages loaded with glycosphingolipids (“Gaucher cells”); Accumulation of these cells in the bone marrow induces cytopenias and bone complica- tions. In a subclinical way, the impaired macrophage function and immune dysregulation involving multiple immune cell lineages leads to chronic metabolic inflammation that has long term consequences. For example, there is a strong association with other entities in older GD1 patients, such as monoclonal gammopathies of uncertain significance (MGUS), hematological B-line malignancies and other solid neoplasms [6]; these adult patients, as well as their carrier relatives, also have an increased risk to develop Parkinson’s disease [7].
Since the 1990s, the incorporation of enzyme replacement therapy (ERT) involving regular intravenous administration of recombinantly produced GluCer has modified the natural history of GD by decreasing accumulation in the target organs and significantly reducing complica- tions. The underlying inflammatory state in type 1 Gaucher disease (GD1) and type 3 (GD3) is reducing [8]. More recently, additional therapeutic strategies have been developed that obtain similar effects, such as oral substrate reduction therapy (SRT) using inhibitors of glu- cosylceramide synthase [9,10]. Patients amenable to treatment with the oral SRT, eliglustat, require careful assessment of current GD status and CYP2D6 phenotypic activity in order to define their dose and apply personalized medicine. The metabolism of eliglustat, like many drugs, is predominantly via CYP2D6, with a minor contribution from CYP3A4; it is also a substrate of ABCB1 [11]. Hence, assessment of drug-drug in- teractions with concurrent medications, food and parapharmacy sup- plements should be performed as standard clinical practice.
Biomarkers in GD, such as chitotriosidase (an activated macrophage marker), the cytokine CCL18/PARC and glucosylsphingosine (the accumulated glycolipid in GD), are important indicators when assessing severity of disease at diagnosis, and are also helpful in monitoring treatment response [12]. Currently, several different treatment options are available for GD, emphasizing the importance of choosing the appropriate individualized therapy for each patient’s needs in order to achieve their therapeutic goals.
Most drugs are metabolized in the liver via metabolic pathways involving the cytochrome P450 (CYP) family of enzymes [13]. CYP2D6 is responsible for metabolizing approXimately 25% of all available medications [14]. The CYP3A sub-family, which includes CYP3A4 as the major liver enzyme, is involved in the metabolism of approXimately 30–50% of all marketed drugs [15]. Variations in the intensity and speed
of drug metabolism, and therefore their effect, are dependent on genetic factors [16]. In the CYP family there are gene polymorphisms respon- sible for variations in metabolic activity, and some of these require analysis prior to the prescription of certain drugs, such as the substrate inhibitor eliglustat [10,17]. It should be noted that proteins other than regulates the transport of molecules across cell membranes by acting as a pump that prevents passage through the blood-brain barrier and re- stricts the absorption of some drugs through the gastrointestinal tract [18]. The sequencing of ABCB1 in different ethnic groups has allowed the identification of the most frequent haplotypes for all three alleles wild type (CT/GT/CT). People with the TTT haplotype tend to show reduced expression of P-gp and, therefore, a deficit in its transport ac- tivity that is not compensated by the normal alleles C or G [19].
In 2004, the government of the Aragon community recognized the FEETEG Lysosomal Storage Unit as a unit of reference; this move allowed many patients from all over the country to receive regular and second opinion visits. All of the patients included in the SGDR have signed their informed consent to become part of the registry and have permitted the use of their data for scientific and academic projects. This study has been performed according with the Helsinki Declaration of 1975 and with the approvals from the Arago´n Ethical Committee (CEICA) and scientific and ethic committees of the FEETEG foundation.
2.2. Patient selection
This is a retrospective-prospective study that includes a cohort of 61 GD1 patients with enzymatically and genetically confirmed diagnoses who are followed in our clinic. The clinical records of these patients were reviewed and analyzed. Data were registered regarding: general demographic characteristics, time from diagnosis to therapy, kind of therapy, type of enzyme, concomitant processes or comorbidities, other drugs taken and adverse effects, clinical and genetic data, degree of disease effect according to the GD-DS3 score [20] and status of disease at the moment of study, including information on spleen status, inflam- matory cytokines and biomarkers, and the GD goals achievement as defined by Pastores et al. [21].
3. Methods
3.1. The study of polymorphisms of CYP2D6, CYP3A4 and three variants of the ABCB1 carrier were performed via TAG®CYP2D6v3, Quantitative
PCR (qPCR) and Restriction Frame Length Polymorphism-Polymerase Chain Reaction (RFLP-PCR)
CYP2D6 genotyping: The allele identification of the CYP2D6 gene (GeneBank database reference: NG_008376.3) was performed according to the TAG®CYP2D6v3 (Luminex Corporation, Texas, USA) manufacturer’s instructions [22]. All reactions included a negative control and a
control sample for each metabolizer status. All patients with gene duplication were analyzed twice.
The aim of this study is to evaluate the most frequent metabolizer profile among GD1 patients with regard to the cytochrome pathway, CYP2D6 and CYP3A4, and ABCB1 transporter, and to consider the importance of this profile in individual applications of eliglustat ther- apy, taking into account patient comorbidities and drug-drug interactions.
2. Patients and methods
2.1. Spanish Gaucher disease registry
The Spanish Foundation for the Study and Therapy of Gaucher Dis- ease (FEETEG) coordinates the SGDR, which has been in place since 1993; more than 420 patients from 120 hospitals have been registered.
3.2. Statistical analysis
Continuous variables analyzed were as follows: age at diagnosis; time of follow-up since inclusion in the SGDR; majority of cases coincide with the date of diagnosis until the last clinical review and time from diagnosis to therapy. To estimate correlations, Pearson’s, Chi-Square, and Mann-Whitney tests were used.
In order to determine the weight of each variable included in the analysis on the response to treatment, a bivariate multiple regression analysis was performed to predict the behaviors of the following vari- ables: age, sex, cytochrome, transporter, treatment, biomarkers, response to treatment, concomitant medications and adverse effects. A partial correlation was also made in relation to response to treatment.
The degree of correlation between variables was classified as follows: 0–0.25, no correlation; 0.25–0.5, weak correlation; 0.5–0.75, moderate correlation and >0.75, strong correlation.
4. Results
4.1. General data
In this study, we have analyzed a cohort of 61 patients (33 females and 28 males, mean age 47.9 years, range: 15–83) diagnosed as having GD1 between ages 6 months to 87 years (mean age 28.1 y). The general patient characteristics are shown in Table 1.
4.2. CYP and P-gp profile
We have classified the patients in order of the frequency of the different types of metabolizers according to the analyzed cytochromes and transporter. The most frequent metabolizer profile in the CYP2D6 group was extensive (49, 80.3%) followed by intermediate (8, 13.1%). For CYP3A4, the most frequent metabolizer profile was extensive for both CYP3A4*1B (90%) and CYP3A4*22 (95%) while normal activity was seen for the three ABCB1 transporter variants in 33 patients (54.0%). The CYP2D6, CYP3A4 and ABCB1 genotype distributions are represented in Fig. 1.
Related to the most frequent metabolizer phenotype (33 patients, 54.1%) were the extensive metabolizer for CYP2D6, CYP3A4*1B and CYP3A4*22 and a normal activity profile for the three ABCB1 SNPs. To a lesser extent, the intermediate metabolizers for CYP2D6, extensive metabolizers for CYP3A4*1B and CYP3A4*22 and normal activity for the three ABCB1 SNPs (6 patients, 9.8%) were related, as were those that are extensive metabolizers for CYP2D6, CYP3A4*1B and CYP3A4*22 and with deficient activity for the three ABCB1 SNPs (4 patients, 6.5%). The remaining phenotypes only include 1 or 2 patients (Table 2).
4.3. Comorbidities and concomitant therapies
Relating to comorbidities, 42 patients (68.8%) had any associated disease. Pathologies affecting the musculoskeletal system (joint replacement, discopathies, osteopenia and fractures) demonstrated the highest percentages, followed by cholelithiasis, repeat infections, arte- rial hypertension and cancer; all of these comorbidities are detailed in Table 3.
More than 54.5% of patients were taking concomitant treatments, most received two or more drugs while only seven patients could be considered as polymedicated (taking more than four different drugs in addition to their specific therapy for GD).Table 4 details the different drugs used in this group of patients, as well as their interactions with the metabolic pathways analyzed.
Regarding response to treatment and metabolizer phenotype, the percentage of patients achieving therapeutic goals was 18.2% in the group of patients not receiving any additional drugs, compared with 54.5% of patients in the group receiving concomitant treatment; these differences had statistical significance (p 0.01).
The incidence of adverse effects was lower (3.3%) in the group of patients not receiving any additional medication when compared to the group taking some type of concomitant drug (14.8%); this difference had statistical significance (p = 0.003).
4.4. Patients under eliglustat therapy
Considering the subgroup of patients under eliglustat therapy (17), the distribution of CYP2D6, CYP3A4 and ABCB1 genotypes was similar to that found in the general group. Ten (58.8%) of these patients have a frequent metabolizer profile and seven (41.2%) of them reached thera- peutic goals with a low incidence of adverse effects. Only one patient had gastrointestinal disorders while, among the seven patients with a less frequent metabolizer profile, the same percentage of responses was observed (5, 29.4%), indicating that the metabolizer profile is not influencing the response. Related to adverse effects, only two patients in this group reported headaches. Considering concomitant therapy, 10 patients were taking two or more drugs on a regular basis and, in 9 of these cases, one or more of the drugs is known to interact with the cy- tochrome and/or ABCB1 transporter pathways analyzed. Most of the drugs being taken were analgesics and NSAIDs; five patients did not reach therapeutic goals while three presented mild or moderate adverse effects, most frequently headache and gastrointestinal disturbances.
Fig. 1. Results of CYP genotype distributions. A) Metabolizer profile of CYP2D6; B) Metabolizer profile of CYP3A4*1B and CYP3A4*22 alleles; C) Genotype results of three SNPs of ABCB1 gene.
4.5. Correlation study
Applying the partial correlation study there is a strong correlation between age and adverse effects (0.098) and weak among CYP3A4*1B and biomarkers ChT (0.301) and LysoGb1 (0.481).The three ABCB1 SNPs are inherited as a haplotype, patients with the TTT haplotype have reduced expression of protein levels and therefore a deficit in transport activity, in this study the frequency of TTT was (6.8%). The individual characteristics of these cases are detailed in Table 5.
5. Discussion
Most drugs are metabolized through the cytochrome pathway com- plex of 18 families. This pathway is also used by dietary foods and nutritional supplements [26], although there are other related factors, such as specific genetic variants, which can influence the response to dietary components and nutrient requirements [27]. The most common pathways are CYP2D6, which is used by 25% of drugs, CYP3A4, which is used by more than 50%, and ABCB1 transporters [28]. Both, drug such as diabetes, hypertension, anxiety and depression, in addition to the musculoskeletal complications they usually suffer [4–6,29]. In this cohort, 17 patients were treated with the substrate inhibitor eliglustat, the dose of which is dependent on the patients’ CYP2D6 metabolizer profile and adjustment for concomitant treatments as may be required [17].
In reference to the distribution of the metabolizer phenotypes found in this study, in the CYP2D6 analysis we observed a higher prevalence of the *1 and *4 alleles compared to the European population [30], as well as of the *1, *2 and *4 alleles compared to the Iberian population, which matches the study carried out on the Spanish Gaucher population [31, 32]. For the *1B and *22 alleles of CYP3A4 and the three SNPs of ABCB1, the prevalence found is similar to that of the Iberian and European populations [33,34].
The most frequent metabolizer phenotype found in this patient cohort is fast for the two types of cytochrome, CYP2D6 and CYP3A4*1B and *22, and normal activity for the three SNPs of the ABCB1 gene. These characteristics do not appear to be correlated with sex or with the variants in the GBA gene of the patients, but a correlation does exist between age, the three SNPs of ABCB1 and the appearance of adverse effects.
In the overall analysis of the patients in the study, we observed a higher percentage of objective responses in the frequent metabolizer profile group compared to the patients with other phenotypes (69% vs 52.4%, p 0.001). Considering the subgroup of patients under treat- ment with eliglustat, for which determining the CYP2D6 phenotype is mandatory in order to establish the indication and posology of the drug (according to the technical data sheet), when classifying them according to metabolizer profile we observed that 10 of them have the most frequent metabolizer profile (10 vs 7). However, the percentage of objective responses obtained is similar in both types (70% vs 71.4%), indicating that the metabolizer profile does not directly influence the response.
Regarding the influence of metabolizer profile on adverse effects, it was observed that siX patients with the most frequent metabolizer pro- file (17.6%) presented mild or moderate adverse effects, while another siX patients in the group with less frequent metabolizer profiles (23.1%) also presented mild or moderate adverse effects (p 0.002). The nine patients who were taking other drugs in their daily lives had a higher incidence of adverse effects compared to patients who did not need concomitant medication (42.6% vs 14.8%, p 0.003). Two patients from the less frequent metabolizer group referenced headaches as an adverse effect; patient 19 was a poor metabolizer for CYP2D6, extensive metabolizer for CYP3A4 and deficient for the ABCB1 gene SNPs rs1045642 and 2032582, while patient 27 was a medium metabolizer for CYP2D6, extensive for CYP3A4 and normal for the 3 SNPs of the ABCB1 gene. This adverse effect can be considered to be of mild rele- vance, however, it leads patients to take analgesics such as paracetamol and fentanyl, both of which are metabolized by the CYP3A4 pathway. We reasoned that, through analyzing the genotype-phenotype relation- ship of patients with GD, we could help in treatment selection as a guide to the application of personalized and precise therapy.
In this series, it was found that more than 54.5% of the patients were taking other therapies in addition to the specific treatment for GD. This percentage is slightly higher than the value registered in the general Spanish population (42% of those over 18 years old take medication very or moderately frequently) according to the survey carried out by Win World Survey (WWS) in 2019 [35].
There is a strong correlation between age and adverse effects, which may be justified in that older patients have more comorbidities and a greater number of concomitant treatments. There is also a weak correlation between CYP3A4*1B and the biomarkers ChT and LysoGb1, for which there is no known explanation so far.
It is also known that people over 65 years of age are the most poly- medicated subjects, taking more than five different drugs a day on average. In this study, polymedicated patients have a lower average age: 50.4 years with a range of 40–62 years old. ParadoXically, the percentage of patients who achieved objective responses is significantly higher among patients taking concomitant medication (54.5% vs 18.2%), regardless of metabolizer phenotype; this is an indirect indicator of the good functionality of their metabolic pathways.