Thiopurine Methyltransferase Levels and Azathioprine Outcomes in Thai Patients with Cutaneous Autoimmune Diseases

Introduction

Thiopurine methyltransferase (TPMT) is a cytosolic enzyme expressed in most cells of the body,¹ with the lowest levels in the brain and lungs and the highest levels in the liver.² The TPMT gene, located on chromosome 6p22.3, is 34 kb in size and comprises 10 exons and nine introns.² TPMT activity is associated with a genetic polymorphism, which is inherited as an autosomal codominant trait. Wild-type homozygous individuals show a higher/normal TPMT activity (the allele is designated as TPMT*1) compared with heterozygous individuals (~10%), whereas individuals with two mutant alleles do not show detectable enzyme activity (~0.3%).^1–3 Approximately 30 variants or mutant alleles of TPMT have been identified, the majority of which demonstrate lower TPMT enzymatic activity or protein expression—TPMT*2, TPMT*3A, TPMT*3B, and TPMT*3C, the four most common mutant alleles, account for approximately 80–95% of individuals with lower TPMT activity.⁴ TPMT*3A and TPMT*3C are the most common variant found in Caucasian (5% frequency) and Asian (2% frequency) populations, respectively.⁵

TPMT catalyzes the S-methylation of aromatic and heterocyclic sulfhydryl compounds, including thiopurine drugs such as the immunosuppressant azathioprine and the anticancer agents 6-mercaptopurine (6-MP) and 6-thioguanine.¹ These drugs have been implicated in several autoimmune diseases, acute lymphoblastic leukemia (ALL), and prevent transplant organ rejection.³ However, despite their wide range of applications, thiopurine drugs yield a relatively narrow therapeutic index, with little difference between therapeutic and toxic doses. Reportedly, high TPMT activity produces more 6-MP but less 6-thioguanine nucleotide (6-tGN), leading to potential treatment failure,³ whereas a lack of TPMT activity increases the concentration of 6-tGN owing to higher substrate accumulation via the hypoxanthine-guanine phosphoribosyltransferase pathway, increasing the risk of adverse effects (AEs), suggesting that the status of TPMT activity is important for the appropriate dose adjustment of thiopurine drugs. Thiopurine drugs cause several AEs, including gastrointestinal, hepatic, and bone marrow toxicity, in nearly 15–28% of patients who discontinue treatment. Among these, bone marrow toxicity is unpredictable; however, it usually develops as an early event, occurs in approximately 2–7% of patients, and may lead to death.⁶ Therefore, the standard of care for monitoring patients treated with azathioprine includes frequent complete blood cell counts and liver function tests and measuring TPMT activity before treatment to prevent severe AEs.⁷ However, to date, no guidelines are available for evaluating TPMT in patients in Thailand. In addition to TPMT, nucleoside diphosphate-linked moiety X-type motif 15 (NUDT15) has emerged as an important pharmacogenetic marker associated with thiopurine-induced toxicity, particularly in Asian populations. Studies have shown that erythrocyte TPMT activity is correlated with other tissues, including the liver and kidney.² To measure TPMT activity, whole blood or erythrocyte lysate is incubated with either 6-MP or 6-tGN, and the methyl donor S-adenosylmethionine, which yields 6-methylmercaptopurine (6-MMP) or 6-methylthioguanine. These reaction products are subsequently measured by different methods, including radio-enzymatic assay, high-performance liquid chromatography (HPLC), and enzyme-linked immunosorbent assay (ELISA).³ Of these methods, ELISA utilizes an antibody specific to the riboside of 6-MMP and provides more accurate results. Furthermore, it is less expensive, simple, and safe.³

The aim of this study was to assess the levels of TPMT enzyme by ELISA in 300 Thai patients diagnosed with various cutaneous autoimmune diseases. Further, the proportion of patients with low levels of TPMT, increased risk of AEs in patients treated with azathioprine, and relationship between TPMT levels and clinical response in patients with pemphigus vulgaris (PV) were evaluated.

Materials and Methods

Patients

Three-hundred Thai patients aged ≥18 years with different cutaneous autoimmune conditions attending the Immunology Clinic, Institute of Dermatology, Bangkok, Thailand between June to September 2017 were enrolled in the study. Eligibility criteria for the study included patients who had previously received, were currently receiving, or currently not receiving azathioprine. Patients were excluded if they received blood transfusions within 3 months before the blood sample collection or if they received drugs such as allopurinol, febuxostat, or aminosalicylates together with azathioprine. Demographic data, concomitant drug information of azathioprine use, adverse reaction, and reason for azathioprine termination were recorded. For patients with pemphigus, the records included first pemphigus rash duration, age at onset, severity at first diagnosis, clinical phenotype at first diagnosis, and first intercellular antibody titer.

The study was approved by the Ethics Committee and all patients provided written informed consent before their enrollment in the study.

Sample Collection

Blood samples (10 mL) from each patient were collected in blood collection and Vacutainer tubes containing EDTA as an anticoagulant, transported to the immunological laboratory, and frozen at ˗20 °C until analysis. For the patients with low TPMT levels, subsequent polymerase chain reaction (PCR) analysis for the TPMT genotype was performed.

Analysis of TPMT Levels

After collection, the whole blood was allowed to clot by leaving it for 15 min and centrifuged at 3000 rpm for 10–15 min. The separated serum (500–1000 µL) was transferred to two microfuge tubes (Eppendorf) and kept at 2–4 °C and −20 °C separately. ELISA kits for assessing human TPMT were obtained from Cusabio Company LLC (Houston, TX, USA). For the determination of TPMT levels, all reagents, working standards, and samples were prepared following the manufacturer’s instructions. Then, 100 µL of each standard and sample was added to each well of a 96-well plate and incubated for 2 h at 37 °C. After incubation, the liquid from each well was replaced with 100 µL of biotin-conjugated antibody (1×), followed by incubation for 1 h at 37 °C. Afterward, all wells were aspirated and washed three times with Wash Buffer Concentrate (25×) and distilled water, followed by the addition of 100 µL of HRP-avidin (1×) to each well and incubation for 1 h at 37 °C. Subsequently, all wells were aspirated and washed five times, and 90 µL of 3,3ʹ,5,5ʹ-tetramethylbenzidine substrate was added to each well and incubated for 15–30 min at 37 °C. Finally, 50 µL of stop solution was added to each well and the absorbance was measured within 5 min using a microplate reader (MPSCREEN Microplate Reader MR-96A (Singapore, Republic of Singapore)) at 450 nm. The standard curve was employed to calculate the results.

Relationship Between TPMT Levels and Clinical Responses

To assess the relationship between TPMT levels and clinical response, 65 patients receiving the combined treatment of prednisolone and azathioprine for at least 1 year were also enrolled in the study. The standard treatment regimen for PV in our setting started with 1 mg/kg/d prednisolone and 1–2 mg/kg/d of azathioprine, which was gradually reduced according to the patient’s clinical response. Clinical response was considered favorable if the patient had no recurrence in the first year of treatment, and the dose of prednisolone did not change by 30 mg/day. In contrast, clinical response was considered unfavorable if the patient had a recurrence with the development of new lesions, and the dose of prednisolone was increased beyond 30 mg/day, or azathioprine was replaced with another immunosuppressive drug or intravenous immunoglobulin (IVIG). Medical records were reviewed by a dermatologist, unaware of the TPMT activity results, to determine clinical responses.

Genomic DNA Extraction

DNA was extracted from blood samples (3–6 mL, mixed with EDTA) collected from the patients with the lowest levels of TPMT enzyme using the MagNA Pure Compact instrument (Roche Applied Science, Pennsburg, Germany). The quality of genomic DNA was assessed using the NanoDropND-1000 (Wilmington, DE, USA). Subsequently, genomic DNA was stored at 2–8 °C for up to 1 week and frozen at −20 °C for 1 month.

TPMT*3A, TPMT*3B, and TPMT*3C Genotyping Assay

TPMT*3A (rs1800460 and rs1142345), TPMT*3B (rs1800460 assay ID: C_27102431_00), and TPMT*3C (rs1142345 assay ID: C19567_20) were evaluated by TaqMan-based analysis (ViiA™7 Real-Time PCR System, Applied Biosystems, Foster City, CA, USA) in a 96-well plate following the manufacturer’s instructions. Subsequently, genotyping was performed in 20 µL containing 5 ng/µL genomic DNA, 10 µL of TaqMan SNP genotyping master mix (Life Technologies), 1 µL of TaqMan assay, and 5 µL of distilled H₂O. The PCR conditions were as follows: initial denaturation at 95 °C for 10 min, followed by 50 cycles of 92 °C for 15s and 60 °C for 90s. Genotyping was interpreted using the algorithm and software supplied by the manufacturer.

Statistical Analyses

The sample size was calculated following the method described by Chisick et al⁸ who showed that <1% of the population has low levels of TPMT. Three-hundred patients were enrolled to detect the prevalence of low TPMT with a 95% confidence level using sample size calculation for rare events.⁹

Quantitative variables are presented as mean and standard deviation (SD), whereas categorical variables are presented as counts and percentages. Student’s t-test and one-way analysis of variance were used to verify the significant differences in means between and among groups, respectively. Fisher’s exact test was used to determine the significant differences in proportions between or among groups because few patients had unfavorable clinical responses. Shapiro–Wilk test was used to determine whether the distribution of TPMT levels was normal. Univariate linear regressions were used to determine the association of any variable with TPMT levels. P <0.05 was considered statistically significant. All statistical analyses and plotting of graphs were performed using R (version 3.4.3, R Foundation for Statistical Computing, Vienna, Austria).

Results

Demographics

Among 300 patients, the mean age was 50 years (SD = 15.3 years), and 203 patients (67.7%) were women (Table 1); 10 (3.3%) smoked, 59 (19.7%) consumed alcohol occasionally, two (0.7%) consumed alcohol frequently, and 42 (14%) reported drug allergy.

Table 1 Demographics and Medical History of Patients Recruited in This Study

Patients had the following cutaneous autoimmune diseases: PV (93, 31%), lupus erythematosus (93, 31%), pemphigus foliaceus (39, 13%), bullous pemphigoid (32, 10.7%), systemic sclerosis (26, 8.6%), dermatomyositis (2, 0.7%), bullous systemic lupus erythematosus (1, 0.3%), CREST syndrome (1, 0.3%), epidermolysis bullosa aquisita (2, 0.7%), LE/LP overlap (3, 1.0%), linear IgA bullous dermatosis (1, 0.3%), mixed connective tissue disease (2, 0.7%), overlapping syndrome (2, 0.7%), pemphigus erythematosus (1, 0.3%), and pemphigus vegetans (2, 0.7%).

TPMT Level

The TPMT level was 6.35–159.8 mU/mL (mean, 84.9 ± 30.5 mU/mL). One patient had an undetectable TPMT level. Although mean TPMT levels varied among diseases, there was no significant difference (p = 0.850). The overall TPMT level distribution did not follow a normal distribution (p = 0.029). However, separate analysis of each disease group revealed that, except for the pemphigus foliaceus group (p = 0.029), TPMT levels in the remaining groups, including the PV (p = 0.592), bullous pemphigoid (p = 0.239), lupus erythematosus (p = 0.160), and systemic sclerosis (p = 0.668) groups, followed normal distribution.

Univariate linear regression analyses revealed a significant correlation between alcohol consumption and lower TPMT levels (p = 0.005), whereas age, sex, smoking, history of drug allergy, and concurrent drug use (such as prednisolone, azathioprine, cyclophosphamide, nicotinic, and tetracycline) were not correlated with TPMT levels (Table 2).

Table 2 Association Between TPMT Levels and Variables Collected in the Study

Effects of Low Levels of TPMT in Development of AEs

AEs leading to azathioprine discontinuation occurred in 17 of 91 patients (18.7%) treated with azathioprine (Table 3). These AEs included hepatitis in 10 patients and severe nausea/vomiting and mild leukopenia in two patients each. Other reasons for azathioprine discontinuation that did not correlate with the given drug included HBV hepatitis in two patients and severe sepsis in one patient. None of the 17 patients experiencing AEs showed low or undetectable TPMT levels. Further, TPMT levels did not correlate with the development of AEs (p = 0.184). Only one patient with an undetectable TPMT level had not received azathioprine therapy.

Table 3 Details of 17 Patients Who Discontinued Azathioprine Therapy Because of Adverse Effects and TPMT Levels

Relationship Between TPMT Levels and Clinical Responses

The relationship between TPMT level and azathioprine efficacy was evaluated in 65 patients with PV and pemphigus foliaceus who were treated for at least 1 year with azathioprine and prednisolone combination therapy. Fifty-seven patients showed favorable, whereas eight patients showed unfavorable, clinical responses (Table 4). Among the 57 patients showing favorable clinical responses, in 32 (56.1%), the prednisolone dose was not increased between azathioprine doses, whereas in the remaining 25 (43.9%), it was increased up to a maximum limit of 30 mg/day throughout the year. Among eight patients with unfavorable clinical responses, the dose of prednisolone was increased beyond 30 mg/day in four. Additionally, azathioprine was substituted in one patient with another immunosuppressive drug, two patients received IVIG, and one patient had both. Table 5 shows the clinical responses with various parameters, such as age, sex, age at the time of disease diagnosis, duration of disease, clinical phenotype, and TPMT level. The characteristics of the two patient groups were not statistically different, except for age of disease onset. Patients with unfavorable clinical responses had a younger disease onset than those with favorable responses (p = 0.013). Moreover, TPMT level did not correlate with clinical responses (p = 0.363).

Table 4 Clinical Response of Patients Receiving Azathioprine Therapy

Table 5 Clinical Response in Patients with Pemphigus Vulgaris and Pemphigus Foliaceus Treated with a Combination of Azathioprine and Prednisolone for One Year and Its Relation to TPMT Level

TPMT Genotype

TPMT genotype analysis performed by real-time PCR using the samples from the 10 patients with low TPMT levels, including one patient with an undetectable level of TPMT, could not identify a mutation of the TPMT gene (Table 6), which could be owing to the small population size in this study.^10,11 The patient with an undetectable TPMT level was 56 years old from Ayutthaya Province. He was diagnosed with pemphigus foliaceus in 2010 by skin biopsy and direct immunofluorescence assay. At that time, his intercellular antibody titer (IIF titer) was 1:160. In 2013, he received prednisolone to control the disease without another immunosuppressive drug and stopped prednisolone in 2015; at that moment, his IIF titer was negative, suggesting that his lesion could be controlled by topical treatment.

Table 6 Genotype of 10 Patients with the Lowest TPMT Levels

Discussion

Azathioprine is widely used to treat cutaneous autoimmune diseases, especially PV, in which the drug is considered a first-line adjuvant immunosuppressant per the European Dermatology Forum guidelines.¹² Although the main advantage of azathioprine is its effective steroid-sparing properties, the drug has been implicated in several AEs, such as nausea, vomiting, pancreatitis, hepatitis, and myelosuppression.¹³

To date, only a handful of studies have reported TPMT genetic variation using PCR analysis in Thai populations. These studies included patients with leukemia, renal transplantation, and systemic lupus erythematosus, who received thiopurine drugs for their treatment, and demonstrated that TPMT*3C is the only variant TPMT allele in these patients. Further, in three patients with myelosuppression, the heterozygous TPMT*1/*3C variant was detected via PCR analysis.^14–18 A similar study that assessed TPMT polymorphism in 200 healthy Thai participants reported TPMT*1/*3C in 18 subjects (9%) and TPMT*3C/*3C in one subject (0.5%).¹⁹ Few studies in Thailand have investigated TPMT activity. Boonsrirat et al¹⁶ by HPLC analysis of TPMT activity, reported an intermediate low phenotype in a patient with systemic lupus erythematosus who suffered from severe myelosuppression after receiving azathioprine. Vannaprasaht et al¹⁸ studied TPMT activity through radioactivity in 139 renal transplant patients who received azathioprine and demonstrated that mean TPMT activity in nine patients who carried the heterozygous TPMT*1/*3C genotype was lower than that in patients carrying the homozygous wild-type genotype. Recently, Wiwattanakul et al²⁰ used liquid chromatography-tandem mass spectrometry (LC-MS/MS) to evaluate TPMT activity in 132 Thai children and found that 10 children (7.5%) had low TPMT activity using a cut-off value of 40 nmol 6-MMP/g Hb/h. Furthermore, assessment of TPMT levels using ELISA kits improves the speed and technical simplicity of TPMT assays.^10,21 Therefore, in this study, we used a TPMT ELISA kit to analyze TPMT levels in 300 adult patients diagnosed with different cutaneous autoimmune diseases.

The prevalence of cutaneous autoimmune diseases was higher in female patients than in male patients, consistent with the general theory of autoimmune diseases.¹⁶ In terms of systemic drug treatment, prednisolone is most used in cutaneous autoimmune diseases, followed by azathioprine, cyclophosphamide, nicotinic, and tetracycline. The most prevalent diseases in cutaneous autoimmune are PV and lupus erythematosus, followed by pemphigus foliaceus, bullous pemphigoid, and systemic sclerosis. However, there was no significant difference in TPMT levels among these five cutaneous autoimmune diseases. The distribution of TPMT levels did not follow a normal distribution, which could be attributed to the distribution of the prevalent autoimmune diseases.

Interestingly, one of the key factors associated with lower TPMT levels in this study was alcohol consumption. However, different studies have reported inconsistent findings, possibly because some of them were conducted in children with leukemia, while others were conducted in adult transplant patients.^22,23 Nevertheless, the results of this study may be reliable because the TPMT enzyme is localized mainly in liver.² However, this finding should be interpreted with caution due to the relatively small number of patients reporting alcohol consumption.

Our results suggested that TPMT levels do not correlate with the development of AEs, in accordance with the results of Firooz et al²⁴ Moreover, azathioprine-induced toxicity is observed in patients with normal TPMT activity, particularly in the Asian population. For instance, a case study has reported that a variant of the nucleoside diphosphate–linked moiety X-type motif (NUDT15) is an important predictor of azathioprine-induced toxicity in patients with normal TPMT activity.^25,26 NUDT15, a member of the Nudix hydrolase family, prevents the incorporation of thiopurine active metabolites thioguanosine triphosphate and thio-deoxyguanosine triphosphate into DNA by their dephosphorylation. It is associated with 6-MP metabolism, which could explain the strong relationship of SNPs in NUDT15 with myelosuppression.^27,28 In an ALL cohort, NUDT15 c.415C>T (p.Arg139Cys) was commonly observed in East Asian and Hispanic subjects, but rarely in European subjects.²² Furthermore, Chiengthong et al²³ reported that among 82 Thai children with ALL, 2.4% were homozygous whereas 12.19% were heterozygous for the mutation c.415C > T. In their study, 6-MP average dosage in heterozygous or homozygous groups of NUDT15 accounted for 80.3%, 61.5%, and 61.1% of the children, respectively, after 2, 4, and 6 months of 6-MP treatment compared with that in the wild-type allele group. Moreover, several studies have found that NUDT15 polymorphisms provide better results than TPMT in predicting the risk for thiopurine-induced leukopenia.^29,30 These results are consistent with international studies indicating that TPMT activity alone may not fully predict thiopurine-induced toxicity, particularly in Asian populations, and further highlight the growing evidence that NUDT15 polymorphism may serve as a more clinically relevant predictor.

Considering these studies, it can be inferred that the 17 patients who discontinued azathioprine owing to AEs and showed normal TPMT levels are carriers of the NUDT15 variant. However, further studies are required to ascertain the association of NUDT15 in patients with normal TPMT levels experiencing thiopurine-induced toxicity.

Patients with PV and pemphigus foliaceus were classified into favorable and unfavorable clinical response groups. Patients with unfavorable clinical responses had a younger age of disease onset than those in the favorable responses group. However, the TPMT level did not correlate with clinical responses. From these findings, it can be inferred that the assessment of the TPMT level might not aid in predicting the clinical responses.

TPMT genotype tests performed by real-time PCR revealed TPMT*1/*1 (wild-type) in all 10 patients with the lowest levels of TPMT, including one patient with an undetectable TPMT level. Genotyping was performed only in patients with the lowest TPMT levels to focus on individuals at highest risk of enzyme deficiency; however, this may limit comprehensive genotype–phenotype correlation across the entire cohort. However, undetected TPMT level in this particular patient might not have resulted from the low TPMT level; rather, it can be explained by several hypotheses, such as the effect of some other factors leading to undetectable TPMT enzyme. Moreover, the small sample size of this study might have failed to detect the rare genotype, and the sensitivity of the TPMT ELISA kit might not be sufficiently high to detect an extremely low enzyme level in this patient. In addition, TPMT levels were measured using ELISA rather than enzyme activity assays, which may limit direct comparison with studies using alternative methodologies. To overcome these limitations, further studies using larger populations are required to compare the ELISA method with LC to assess its efficiency in accurately detecting TPMT enzyme levels and activities and validating a reliable method when using PCR for TPMT genotyping. These findings have important clinical implications for dermatologists prescribing azathioprine. In settings where TPMT testing is not routinely available or does not predict adverse effects, careful clinical monitoring and consideration of alternative pharmacogenetic markers such as NUDT15 may be warranted.

In summary, this study reports the levels of TPMT enzyme in Thai patients with varied cutaneous autoimmune diseases. The analysis of polymorphism in the TPMT gene in these patients revealed TPMT*1/*1 (wild-type) in patients with the lower levels of TPMT. The findings demonstrate that the TPMT level did not correlate with the development of AEs, and the assessment of the TPMT levels might not aid in predicting the clinical responses, whereas the patients experiencing AEs with normal TPMT levels could be the carriers of the NUDT15 variant. Collectively, these findings suggest that the efficiency of TPMT levels in predicting the clinical responses to azathioprine therapy could only be sufficiently explored using a large population. Additionally, these findings highlight the potential role of NUDT15 polymorphism as a more clinically relevant alternative predictor of thiopurine toxicity, particularly in Asian populations.