The influence of serial 50 μL microsampling on rats administered azathioprine, the immunosuppressive drug

Abstract

According to the ICH S3A Q&A, microsampling is applicable to pharmaceutical drugs and toxicological analysis. Few studies have reported the effect of microsampling on the toxicity of immunotoxicological drugs. The aim of this multicenter study was to evaluate the toxicological effects of serial microsampling on rats treated with azathioprine as a model drug with immunotoxic effects. Fifty microliters of blood were collected from the jugular vein of Sprague-Dawley rats at six time points from day 1 to 2 and 7 time points from day 27 to 28. The study was performed at three organizations independently. The microsampling effect on clinical signs, body weights, food consumption, hematological parameters, biochemical parameters, urinary parameters, organ weights, and tissue pathology was evaluated. Azathioprine-induced changes were observed in certain hematological and biochemical parameters and thymus weight and pathology. Microsampling produced minimal or no effects on almost all parameters; however, at 2 organizations, azathioprine-induced changes were apparently masked for two leukocytic, one coagulation, and two biochemical parameters. In conclusion, azathioprine toxicity could be assessed appropriately as overall profiles even with blood microsampling. However, microsampling may influence azathioprine-induced changes in certain parameters, especially leukocytic parameters, and its usage should be carefully considered.

Graphical Abstract

Highlights

• •The influence of 50 μL microsampling was assessed in azathioprine-treated rats.
• •Microsampling did not affect most toxicity parameters altered by azathioprine.
• •Microsampling may mask changes in leukocytic and biochemical parameters.

1. Introduction

Toxicokinetic (TK) analysis is performed through non-clinical safety. In rat studies, repeated blood sampling is required; 100–200 μL of blood is collected at each point. To avoid the influence of blood sampling from the same animals on toxicological profiles, satellite animals are usually used for TK analysis. Owing to recent developments in analytical technology, drug concentrations can be determined from small amounts of samples by highly sensitive analytical instruments. Microsampling can be used to collect less than or equal to 50 μL of blood at each point, and the relationship between toxicological and TK parameters can be evaluated in each animal. Considering the 3Rs in animal research, a decrease in the number of study animals and sampling volumes contributes towards refinement and reduction. ICH S3A Q&A, which focused on microsampling, was released for application in drug development studies [1].

According to the ICH S3A Q&A, microsampling can be applied to the toxicological analysis of pharmaceutical drugs including biopharmaceuticals, if the drug concentration is measured with the same sensitivity as with a conventional sampling method (Q3). When evaluating the influence of sequential sampling, it is important to consider changes in physiological conditions. If previous studies show that test drug-related changes to hematological parameters could be exacerbated by frequent blood sampling, or it is suspected that the pharmacological action of the test drug may induce such effects, the use of satellite groups of animals for TK assessment would be warranted, even if microsampling techniques are used (Q6).

We have reported that, in a 28-day study in rats, jugular vein microsampling (50 μL) at six timepoints from day 1 to 2 and 7 timepoints from day 27 to 28 did not significantly affect the parameters for toxicological evaluation on non-drug treatment rats [2], [3]. Thus, this microsampling method can be applied in serial blood collection for evaluating both toxicology and TK in each animal. Additionally, we have evaluated the effect of jugular vein microsampling on the toxicological profile of phenacetin, which induces hematological toxicity. Serial microsampling (50 μL) exhibited minimal influence on the assessment of hematological parameters. However, using microsampling, we detected apparent deterioration or masking of phenacetin toxicity with statistical significance with respect to certain sporadic parameters (body weight, plasma ALT activity, urine volume, and liver weight), primarily at one organization. Further studies clearly are needed, because at present, the effects of microsampling on the toxicological profiles of drugs have not been fully assessed, especially for drugs with other toxicological characteristics, especially immunotoxicity. This is because frequent puncture for serial microsampling could be stressful for animals, leading to immunosuppressive effects. The immunosuppressive effect associated with the stress induced by frequent microsampling can affect the assessment of drugs with immunotoxic effects. In case of immunosuppression caused by the cytotoxic effect of the drugs, hematotoxicity should be also observed, which may also be influenced by serial microsampling. In other words, these changes of the hematological parameters might be enhanced or masked by microsampling. Therefore, precise evaluation of the effect of frequent microsampling on the toxicological profile of an immunosuppressive drug is very important. In the present study, at three independent organizations to fully understand the microsampling effect, we evaluated the toxicological effects of serial microsampling on rats treated with azathioprine as a model drug, known to induce immunotoxicity and hematotoxicity. TK measurements were also performed to confirm the dose-dependent exposure of azathioprine on rats and the differences in TK parameters between the three organizations.

2. Materials and methods

2.1. Organizations and animals

The animal experiments were performed at three organizations (A, B, and C). Sprague-Dawley rats (Crl:CD, female, 5 weeks of age, c.a., 165 g of weights) were purchased from the Jackson Laboratory Japan, Inc. (Yokohama, Japan). Female rats were selected because of their possible higher sensitivity to microsampling due to lower circulatory blood volume than that of male rats. The rats are housed in room maintained at 19–26 °C, 30–75 % relative humidity, 12/12 h light-dark cycle, and ventilation rate of 6–20 times/h. They were housed one animal per cage at organization A and C and two or three animals per cage at organization B. These rats were fed CR-LPF (B and C, Oriental Yeast Co., Tokyo, Japan) or CRF-1 (A, Oriental Yeast Co., Tokyo, Japan). Drug administration was started after 5–7 days of habituation at each organization. The study was initiated after approval by the Animal Experiment Committee of each organization.

2.2. Study protocol

Two groups were set in each organization: no-treatment group (group I) and 50 μL-sampling group (group II), comprising 5 female rats each. Based on the data of azathioprine repeated treatment study by Toxicogenomics Project [4], the azathioprine doses were selected as 3 mg/kg for non-observed adverse effect level (NOAEL) dose and 10 mg/kg for lowest observed adverse effect level (LOAEL) dose assessed by reduction of leukocyte and neutrophil counts and thymus atrophy. We administered doses of 0, 3, and 10 mg/kg of azathioprine suspended in 0.5 % methylcellulose aqueous solution daily per oral for 28 days. The animals were randomly (stratified random sampling) allocated to each group based on the body weights measured on the day of allocation so that the initial mean body weights of each group were equivalent. Fifty microliters of blood were collected from the jugular vein of the group II rats at 0.5, 1, 2, 4, 8, and 24 h after dosing on days 1 and 0, 0.5, 1, 2, 4, 8, and 24 h after dosing on day 27. Blood collection was performed using 3/10-mL low-dose syringes with 29 G needles (Becton Dickinson Co., Franklin Lakes, NJ, USA), and heparin sodium was added to the blood. All animals were assessed daily for clinical signs and their body weight and food consumption measured once a week. Urine was collected in week 4.

Rats were fasted from the evening of day 28 overnight and were anesthetized with isoflurane (Mylan V.V., Tokyo, Japan or Pfizer Inc., New York, NY, USA). Blood was collected from the abdominal vena cava for assessment of hematology and biochemistry. Subsequently, the rats were euthanized by exsanguination, and necropsy was conducted. Next, the organs were collected and subject to a histopathology test (essential organs examined were the thymus, spleen, mesenteric lymph node, and submandibular lymph node). The assessed hematological parameters were red blood cell (RBC) count, hematocrit, hemoglobin, mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), reticulocyte count, mean corpuscular volume (MCV), leukocyte/white blood cell (WBC), neutrophils, lymphocytes, monocytes, basophils, eosinophils, platelet counts, fibrinogen, prothrombin time (PT), and activated partial thromboplastin time (APTT) (at maximum). The measured biochemical parameters were total protein, albumin, albumin/globulin (A/G) ratio, α1-globulin, γ-globulin, glucose, total cholesterol, triacylglycerol, phospholipid, alanine transaminase (ALT), aspartate transaminase (AST), glutamate dehydrogenase (GLDH), alkaline phosphatase (ALP), total bilirubin, lactate dehydrogenase (LDH), guanase, γ-glutamyltranspeptidase (γ-GT), creatine phosphokinase (CPK), blood urea nitrogen (BUN), creatinine (Cre), sodium (Na), potassium (K), calcium (Ca), inorganic phosphorus (P), and chloride (Cl) (at maximum). The following urinary parameters were assessed: volume, specific gravity, osmolality, Na, K, and Cl (at maximum).

2.3. TK evaluation of azathioprine

Azathioprine is rapidly catabolized to 6-mercaptopurine (6-MP) non-enzymatically in the blood stream; thus, we evaluated plasma 6-MP concentrations in this TK evaluation. We used a modified sample extraction method reported previously [5], and 6-MP concentrations were measured using a liquid chromatography system with tandem mass spectrometry. For extracting 6-MP, 10 μL of plasma was mixed with 20 μL of 6-MP-¹³C, ¹⁵N₂ isotope (internal standard), and 100 μL of dithiothreitol, following which 8 μL of perchloric acid was added to the mixture and vortexed thoroughly. The mixture was placed on ice for 10 min and centrifuged at 10,000 × g for 10 min at 4 °C. The supernatant (100 μL) was transferred to other tubes and mixed with 100 μL of 0.1 % formic acid. After filtration, 1 μL of mixture was injected into the ultra-high performance liquid chromatography (UHPLC) system. Chromatographic separation was performed using Ultimate 3000 UHPLC system (ThermoFisher Scientific, Waltham, MA USA) with a Cortecs C18 + (1.6 µm, 2.1 × 50 mm) column (Waters, Ireland) maintained at 30 °C. Mobile phase A was composed of 0.1 % formic acid in water, and B was 0.1 % formic acid in acetonitrile. The flow rate was 0.4 mL/min, and the initial composition of the mobile phase was 1 % of mobile phase B. Subsequently, we applied a linear gradient of 3 % of mobile phase B till 1.1 min, increasing the concentration to 30 % at 2.8 min, and decreasing it to 1 % at 3.2 min. Detection of 6-MP was performed using an TSQ Vantage Triple-Stage Quadrupole Mass Spectrometer (ThermoFisher Scientific, Waltham, MA, USA) in positive multiple reaction monitoring mode.

2.4. Statistical analysis

Equal variances were estimated among groups administered 0, 3 and 10 mg/kg azathioprine using the Bartlett test; p < 0.01 was considered to indicate statistical significance, separately for groups with and without microsampling. The group mean differences were estimated by the Dunnett test for equal variance and the Steel test for non-equal variance. Differences in body weight, food consumption, blood parameters, biochemical parameters, urinary parameters, and organ weights were compared. MiTOX Computer System (Mitsui E&S Systems Research Inc.) and tsPharma LabSite, (Fujitsu Limited) were used for statistical analysis. Correction for multiple comparisons was not applied to find even minimal influences of microsampling.

3. Results

3.1. Clinical signs, body weight, and food consumption

No clinical signs were related to treatment with azathioprine or microsampling (data not shown). Body weight and food consumption were assessed twice a week; the total values per week are shown in Supplementary Table 1. The body weights increased from treatment initiation to week 4 in all groups, and no significant difference was observed between the non-treatment group and each azathioprine-treated group at three organizations. The influence of azathioprine on food consumption was slightly different between group I and II at organization B. In group I, azathioprine treatment significantly decreased food consumption in weeks 2 and 3 compared with the vehicle control treatment. In contrast, in group II, significant increases in food consumption were observed in weeks 1, 2, and 4 for the 3 mg/kg dose and in week 1 for the 10 mg/kg dose compared with that for the 0 mg/kg dose. However, these differences in food consumption were small and not detected at the other organizations. These results suggested that 50 μL of sampling had no or minimal influence on the assessment of azathioprine toxicity with respect to clinical signs, body weight, and food consumption.

3.2. Hematology, biochemistry, and urinalysis

Hematological parameters were assessed on day 29 (Table 1). Azathioprine treatment tended to decrease the values of erythroid parameters (RBC counts, hematocrit, and hemoglobin), leukocytic parameters (WBC counts, neutrophil counts, lymphocyte counts, and eosinophil counts), and PT compared with vehicle control (0 mg/kg). In group I, treatment with 10 mg/kg azathioprine decreased the values of the erythroid and leukocytic parameters. At organization B, the azathioprine-treatment group showed a statistically significant decrease in RBC count (10 mg/kg), hematocrit values (10 mg/kg), hemoglobin levels (10 mg/kg), WBC count (10 mg/kg), lymphocyte count (10 mg/kg), eosinophil count, (≥ 3 mg/kg), and monocyte count (≥3 mg/kg) compared with the vehicle control group (0 mg/kg). At organization C, the WBC count and lymphocyte count were significantly decreased in the 10 mg/kg azathioprine-treatment group than in the 0 mg/kg. In group II, almost all hematological parameters showed a similar trend to those in group I; however, significant differences in erythroid and leukocytic parameters were diminished or masked at organizations B and C. These results showed that microsampling contributed to the azathioprine-induced reduction in the values of erythroid parameters. In contrast, at organization C, only group II showed statistically significant differences in erythroid parameters. At organization A, both groups I and II showed no statistical difference in the three erythroid parameters. Thus, a systematic trend was not observed among the organizations, suggesting that these effects of microsampling were less meaningful in terms of the toxicological profile. Systemic trends (at least two organizations showed the same trends in statistical evaluation) for masking of toxicological changes by microsampling were shown with respect to WBC count (organizations B and C), lymphocyte count (organizations B and C), and PT (organizations A and B); the statistical significance observed in group I of the 10 mg/kg azathioprine group was absent in the corresponding group II. Serial 50 μL microsampling apparently masked changes in WBC count and coagulation-related parameters.

Table 1.

Hematological parameters in azathioprine-administered rats treated with 50 μL microsampling and corresponding controls.

		Organization A			Organization B			Organization C
Azathioprine (mg/kg)		0	3	10	0	3	10	0	3	10
RBC	Group I	7.61 ± 0.23	7.52 ± 0.20	7.30 ± 0.25	7.69 ± 0.27	7.64 ± 0.27	7.11 ± 0.21**	7.93 ± 0.29	7.92 ± 0.48	7.37 ± 0.60
(× 106/μL)	Group II	7.30 ± 0.49	7.62 ± 0.28	7.18 ± 0.23	7.24 ± 0.20	7.04 ± 0.38	6.89 ± 0.24	7.44 ± 0.30	7.53 ± 0.19	6.61 ± 0.73*,†
Hematocrit	Group I	42.2 ± 1.5	42.0 ± 1.0	40.9 ± 1.4	42.2 ± 1.0	41.8 ± 1.9	38.9 ± 1.0**	45.1 ± 2.0	43.4 ± 2.4	42.2 ± 2.4
(%)	Group II	40.7 ± 1.9	41.7 ± 1.0	40.1 ± 0.5	40.5 ± 0.3	39.0 ± 1.8	38.1 ± 1.5*,†	42.4 ± 1.7	42.1 ± 0.9	38.6 ± 2.8*,†
Hemoglobin	Group I	15.5 ± 0.5	15.5 ± 0.4	14.9 ± 0.6	15.2 ± 0.3	15.1 ± 0.5	13.9 ± 0.4**	16.2 ± 0.8	15.8 ± 0.9	14.9 ± 1.0
(g/dL)	Group II	15.1 ± 0.6	15.4 ± 0.3	14.7 ± 0.3	14.4 ± 0.2	13.9 ± 0.7	13.5 ± 0.6*,†	15.3 ± 0.5	15.1 ± 0.4	13.4 ± 1.2**,†
MCH	Group I	20.3 ± 0.6†	20.7 ± 0.5†	20.4 ± 0.6†	19.7 ± 0.5	19.7 ± 0.8	19.5 ± 0.5	20.4 ± 0.4	20.0 ± 0.7	20.2 ± 0.6
(pg)	Group II	20.7 ± 0.8†	20.2 ± 0.4	20.6 ± 0.9†	19.9 ± 0.4	19.8 ± 0.3	19.6 ± 0.4	20.5 ± 0.7	20.0 ± 0.3	20.4 ± 0.7
MCHC	Group I	36.7 ± 0.6†	37.0 ± 0.3†	36.3 ± 0.8†	36.0 ± 0.4	36.1 ± 0.4	35.7 ± 0.6	35.9 ± 0.6	36.5 ± 0.5	35.2 ± 0.4
(g/dL)	Group II	37.1 ± 0.5†	36.8 ± 0.3†	36.8 ± 0.7†	35.6 ± 0.4	35.7 ± 0.2	35.4 ± 0.3	35.9 ± 0.8	35.8 ± 0.3	34.8 ± 0.7*
Reticulocyte	Group I	2.10 ± 0.20	2.15 ± 0.43	2.54 ± 0.65	3.80 ± 1.17	3.52 ± 1.15	4.03 ± 0.81	2.4 ± 0.5	2.3 ± 0.2	2.5 ± 0.3
(%)	Group II	2.45 ± 0.50	2.34 ± 0.32	2.67 ± 0.50	4.10 ± 0.86	5.30 ± 1.34†	4.25 ± 1.01	2.3 ± 0.4	2.9 ± 0.5	3.6 ± 1.2†
MCV	Group I	55.4 ± 1.1	55.9 ± 1.3	56.0 ± 1.9	54.9 ± 1.8	54.7 ± 2.6	54.7 ± 1.6	56.8 ± 0.7	54.8 ± 2.1	57.3 ± 1.7
(fL)	Group II	55.8 ± 1.7	54.9 ± 1.2	55.9 ± 1.7	56.0 ± 1.5	55.5 ± 1.0	55.3 ± 0.9	57.1 ± 1.3	55.9 ± 0.6	58.7 ± 2.7
White blood cell	Group I	5.53 ± 1.26	5.80 ± 1.27	4.00 ± 0.83	10.75 ± 1.21	7.75 ± 1.37	6.92 ± 3.11*	7.6 ± 1.6	5.7 ± 2.4	3.9 ± 0.4**
(× 103/μL)	Group II	6.39 ± 1.59	5.94 ± 2.32	4.74 ± 1.51	9.35 ± 1.99	7.71 ± 1.80	6.54 ± 2.39	6.9 ± 2.2	6.0 ± 1.4	4.8 ± 1.8
Neutrophil	Group I	0.62 ± 0.19	0.66 ± 0.31	0.61 ± 0.25	1.10 ± 0.48	0.74 ± 0.41	0.67 ± 0.23	0.85 ± 0.22	0.64 ± 0.29	0.47 ± 0.14
(× 103/μL)	Group II	0.85 ± 0.28	0.82 ± 0.35	0.67 ± 0.21	0.89 ± 0.40	0.91 ± 0.44	0.50 ± 0.23	1.05 ± 0.48	0.84 ± 0.34	0.54 ± 0.20
Lymphocyte	Group I	4.66 ± 1.21	4.85 ± 1.19	3.20 ± 0.64	9.13 ± 1.13	6.71 ± 0.97	5.93 ± 2.95*	6.38 ± 1.61	4.65 ± 2.11	3.22 ± 0.27*
(× 103/μL)	Group II	5.31 ± 1.81	4.83 ± 2.11	3.84 ± 1.26	8.04 ± 1.80	6.44 ± 1.47	5.79 ± 2.08	5.43 ± 1.93	4.83 ± 1.46	3.96 ± 1.51
Monocyte	Group I	0.09 ± 0.04	0.11 ± 0.04	0.09 ± 0.04	0.39 ± 0.07†	0.22 ± 0.08**	0.26 ± 0.09*	0.18 ± 0.07	0.14 ± 0.07	0.11 ± 0.04
(× 103/μL)	Group II	0.09 ± 0.02	0.14 ± 0.09	0.11 ± 0.06	0.28 ± 0.12	0.24 ± 0.12	0.18 ± 0.08	0.16 ± 0.10	0.14 ± 0.05	0.12 ± 0.07
Basophil	Group I	0.00 ± 0.01	0.01 ± 0.00	0.00 ± 0.00	0.00 ± 0.01	0.00 ± 0.00	0.00 ± 0.00	0.02 ± 0.01	0.01 ± 0.00	0.00 ± 0.01
(× 103/μL)	Group II	0.01 ± 0.01	0.00 ± 0.01	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.01 ± 0.00	0.01 ± 0.01	0.00 ± 0.01
Eosinophil	Group I	0.11 ± 0.04	0.14 ± 0.07	0.07 ± 0.03	0.12 ± 0.03	0.07 ± 0.02*	0.07 ± 0.02*	0.14 ± 0.05	0.13 ± 0.05	0.07 ± 0.03
(× 103/μL)	Group II	0.10 ± 0.02	0.12 ± 0.07	0.07 ± 0.03	0.14 ± 0.05	0.13 ± 0.06	0.08 ± 0.04	0.15 ± 0.07	0.14 ± 0.03	0.09 ± 0.03
Platelet	Group I	1.12 ± 0.22	1.16 ± 0.24	1.12 ± 0.06	1.11 ± 0.13	1.18 ± 0.10	1.28 ± 0.17	1.34 ± 0.10	1.21 ± 0.12	1.34 ± 0.07
(× 106/μL)	Group II	1.03 ± 0.07	1.11 ± 0.17	1.06 ± 0.09	1.13 ± 0.14	1.15 ± 0.14	1.24 ± 0.13	1.33 ± 0.13	1.41 ± 0.24	1.43 ± 0.20
Fibrinogen	Group I	180 ± 8	196 ± 8*	198 ± 12*	189 ± 20	196 ± 14	199 ± 11	176 ± 19	186 ± 16	200 ± 24
(mg/dL)	Group II	205 ± 18	202 ± 11	207 ± 24	208 ± 18	229 ± 28†	221 ± 17	207 ± 26	223 ± 27	228 ± 38
Prothrombin time	Group I	10.6 ± 0.3	10.3 ± 0.1	10.1 ± 0.3*	8.7 ± 0.6	8.1 ± 0.2	8.0 ± 0.3*	9.5 ± 0.1	9.4 ± 0.1	9.3 ± 0.1
(PT, s)	Group II	10.3 ± 0.2	10.4 ± 0.4	10.1 ± 0.2	8.4 ± 0.2	8.0 ± 0.3*	8.1 ± 0.2	9.4 ± 0.1	9.3 ± 0.1	9.1 ± 0.2*
Activated partial thromboplastine time	Group I	17.6 ± 1.2	17.4 ± 1.3	16.6 ± 1.7	13.7 ± 0.7	14.2 ± 1.0	13.7 ± 1.2	13.3 ± 0.8	13.3 ± 1.7	12.5 ± 1.4
(APTT, s)	Group II	16.7 ± 1.3	16.8 ± 2.1	16.8 ± 0.4	14.5 ± 0.6	13.8 ± 0.7	14.0 ± 1.0	13.6 ± 1.9	13.1 ± 1.4	12.2 ± 1.2

Group I, no treatment; Group II, microsampling group.

Results were shown in values ± SD.

Significantly different from the corresponding "0 mg/kg" group by Dunett test or Steel test (*p < 0.05, **p < 0.01).

†: The data are outside the range of background values.

–: not evaluated.

The biochemical parameters of blood on day 29 are shown in Supplementary Table 2. Systemic changes (statistically significant changes observed in at least two organizations) were detected in plasma total cholesterol concentrations, which showed dose-dependent elevation upon treatment with 3 or 10 mg/kg azathioprine in group I at all three organizations. The significant elevation in total cholesterol was absent in group II at organization A and B. In addition, the significant elevation of Ca concentrations with 10 mg/kg azathioprine in group I at organizations B and C was absent in group II. Thus, these two systemic changes appeared to have been masked by microsampling. The other parameter changes were detected only at one of the organizations; therefore, they appeared to have occurred by chance. At organization A, a significant reduction in γ-globulin concentrations was observed in group I treated with azathioprine; however, this significant difference was absent in group II. At organization B, the values of total protein (10 mg/kg), albumin and phospholipid (3 and 10 mg/kg), and A/G ratio (3 mg/kg) in group I showed significant differences between the azathioprine treatment group and vehicle control group (0 mg/kg); however, these differences were not observed in group II.

Urinalysis parameters, which were evaluated in week 4, were not significantly different between the vehicle control (0 mg/kg) group and either of the azathioprine treatment (3 and 10 mg/kg) groups at each organization (Supplementary Table 3).

3.3. Organ weight, necropsy, and pathological examinations

The results for organ weight are shown in Table 2 and Supplementary 4. Rats treated with 10 mg/kg azathioprine showed a tendency towards reduction in thymus weight at all organizations, with statistically significant reduction at organizations B (group I and II) and C (group II), compared with the rats treated with the vehicle control (0 mg/kg). At organization C, statistically significant reduction in thymus weight was observed only 10 mg/kg in group II, with apparently additional deterioration by microsampling. On the other hand, at organization A and B, the weights of the thymus tended to be reduced both in group I and II, with statistical significance at organization B, and the percentage of the reduction were similar between groups I and II, suggesting that deteriorating effect by microsampling at organization C was not replicated at the other organizations. Additionally, in organization A, the weights of the heart and kidney with respect to the body weight were significantly increased with 10 mg/kg treatment in groups I and II, and the relative weight of the liver significantly increased with 10 mg/kg treatment in group I. However, these changes were not replicated at the other organizations. These results suggested that the 50 μL sampling had no or minimal influence on the assessment of azathioprine toxicity with respect to organ weight.

Table 2.

Organ weights in azathioprine-administered rats treated with 50 μL microsampling and corresponding controls.

		Organization A			Organization B			Organization C
Azathioprine (mg/kg)		0	3	10	0	3	10	0	3	10
Heart (g)	Group I	0.857 ± 0.062	0.920 ± 0.095	0.927 ± 0.009	0.821 ± 0.039	0.831 ± 0.059	0.799 ± 0.080	0.814 ± 0.055	0.801 ± 0.034	0.800 ± 0.075
	Group II	0.888 ± 0.093	0.877 ± 0.079	0.923 ± 0.121	0.850 ± 0.090	0.839 ± 0.051	0.797 ± 0.019	0.806 ± 0.063	0.817 ± 0.072	0.802 ± 0.032
Lung (g)	Group I	–	–	–	1.13 ± 0.06	1.14 ± 0.08	1.14 ± 0.02	1.03 ± 0.07	1.03 ± 0.05	1.06 ± 0.07
	Group II	–	–	–	1.17 ± 0.04	1.19 ± 0.07	1.15 ± 0.07	1.01 ± 0.11	1.04 ± 0.09	1.09 ± 0.08
Liver (g)	Group I	6.45 ± 0.56	6.82 ± 0.95	7.34 ± 0.98	6.07 ± 0.33	6.33 ± 0.54	6.54 ± 0.60	6.16 ± 0.66	6.09 ± 0.43	6.34 ± 0.72
	Group II	6.66 ± 0.57	6.73 ± 0.54	6.75 ± 1.13	6.33 ± 0.71	6.28 ± 0.47	6.39 ± 0.39	6.04 ± 1.02	6.21 ± 0.12	6.56 ± 0.72
Kidneys (g)	Group I	1.65 ± 0.11	1.79 ± 0.18	1.76 ± 0.18	1.69 ± 0.13	1.70 ± 0.14	1.65 ± 0.19	1.63 ± 0.20	1.62 ± 0.050	1.70 ± 0.15
	Group II	1.64 ± 0.12	1.77 ± 0.19	1.72 ± 0.13	1.80 ± 0.14	1.79 ± 0.12	1.66 ± 0.08	1.68 ± 0.30	1.65 ± 0.06	1.70 ± 0.11
Thymus (g)	Group I	0.447 ± 0.118	0.496 ± 0.152	0.322 ± 0.125	0.416 ± 0.034	0.380 ± 0.092	0.206 ± 0.067**	0.369 ± 0.056	0.353 ± 0.126	0.296 ± 0.185
	Group II	0.472 ± 0.100	0.425 ± 0.097	0.389 ± 0.107	0.485 ± 0.041	0.439 ± 0.079	0.266 ± 0.115**	0.428 ± 0.094	0.435 ± 0.043	0.246 ± 0.165*
Spleen (g)	Group I	0.455 ± 0.061	0.430 ± 0.054	0.438 ± 0.050	0.491 ± 0.058	0.479 ± 0.044	0.437 ± 0.090	0.423 ± 0.052	0.447 ± 0.096	0.426 ± 0.026
	Group II	0.467 ± 0.039	0.516 ± 0.058	0.423 ± 0.063	0.523 ± 0.046	0.475 ± 0.044	0.435 ± 0.055*	0.440 ± 0.081	0.440 ± 0.021	0.448 ± 0.038
Brain (g)	Group I	1.88 ± 0.11	1.86 ± 0.05	1.92 ± 0.04	–	–	–	1.96 ± 0.13	1.98 ± 0.09	1.93 ± 0.10
	Group II	1.86 ± 0.11	1.81 ± 0.09	1.83 ± 0.10	–	–	–	1.93 ± 0.08	1.97 ± 0.09	2.00 ± 0.05
Pituitary (g)	Group I	0.015 ± 0.002	0.014 ± 0.002	0.011 ± 0.002*	–	–	–	0.012 ± 0.001	0.012 ± 0.002	0.012 ± 0.002
	Group II	0.012 ± 0.001	0.010 ± 0.001	0.012 ± 0.003	–	–	–	0.013 ± 0.002	0.012 ± 0.003	0.012 ± 0.002
Submandibular (g)	Group I	0.448 ± 0.034	0.404 ± 0.035	0.429 ± 0.059	–	–	–	0.428 ± 0.040	0.426 ± 0.030	0.416 ± 0.046
	Group II	0.399 ± 0.041	0.430 ± 0.053	0.404 ± 0.041	–	–	–	0.409 ± 0.073	0.392 ± 0.040	0.423 ± 0.031
Uterus (g)	Group I	–	–	–	–	–	–	0.422 ± 0.088	0.410 ± 0.096	0.523 ± 0.264
	Group II	–	–	–	–	–	–	0.390 ± 0.042	0.592 ± 0.409	0.373 ± 0.060
Ovary (g)	Group I	0.088 ± 0.024	0.086 ± 0.016	0.099 ± 0.020	–	–	–	0.094 ± 0.010	0.088 ± 0.012	0.089 ± 0.012
	Group II	0.084 ± 0.019	0.082 ± 0.014	0.091 ± 0.015	–	–	–	0.094 ± 0.012	0.094 ± 0.009	0.091 ± 0.019
Pancreas (g)	Group I	–	–	–	–	–	–	0.817 ± 0.227	0.939 ± 0.194	0.952 ± 0.204
	Group II	–	–	–	–	–	–	0.881 ± 0.137	0.863 ± 0.042	0.702 ± 0.105*
Adrenal glands (mg)	Group I	62 ± 10	60 ± 5	68 ± 13	–	–	–	55 ± 9	57 ± 7	60 ± 10
	Group II	65 ± 13	66 ± 3	61 ± 7	–	–	–	62 ± 10	63 ± 8	60 ± 12

Group I, no treatment; Group II, microsampling group.

Results were shown in values ± SD.

Significantly different from the corresponding "0 mg/kg" group by Dunett test or Steel test (*p < 0.05, **p < 0.01).

–: not evaluated.

In the necropsy test, the black area in the glandular mucosa of stomach was observed in one rat in each group, and small-sized thymus was observed in one rat of group I treated with 10 mg/kg azathioprine at organization A (Supplementary Table 5). Hemorrhage in the subcutaneous tissue due to microsampling was observed in group II at organization A. At organization B, the small-sized thymus was observed in three of five rats treated with 10 mg/kg azathioprine for both groups. At organization C, a small-sized thymus and hydrocephalus brain was observed in one rat, respectively, treated with 10 mg/kg azathioprine in group II. Hemorrhage in the subcutaneous area was observed in one rat treated with 10 mg/kg azathioprine in group II. These observed changes were not systemic but incidental by azathioprine treatment because these were detected only at one organization, mostly for both groups.

On pathological examination, the lymphocyte cellularity in the thymus cortex were decreased in the rats treated with 10 mg/kg azathioprine at all three organizations; no apparent microsampling effect was observed (Supplementary Table 6). At organization A, a brown pigment deposition was frequently observed in the spleen following azathioprine treatment, irrespective of the microsampling. Many other pathological changes such as fatty change of the hepatocyte were sparsely and independently observed for azathioprine treatment, as well as microsampling.

3.4. Toxicokinetics of azathioprine metabolite

TK parameters of the azathioprine metabolite 6-MP in each organization are shown in Supplementary Table 7 and Supplementary Fig. 1. T_max were 0.5 h on the 1st dose and 27th doses. Following 3 mg/kg treatment at the 1st and 27th doses, the average values of C_max (ng/mL) were 53.9 ± 15.4 and 67.9 ± 16.4, and the plasma concentration was not detected after 4 h from the last treatment. The average AUC (ng∙hour/mL) at the 1st and 27th doses of 3 mg/kg were 50.5 ± 7.1 and 56.8 ± 8.3, respectively. Following the 10 mg/kg treatment at the 1st and 27th doses, the average C_max values (ng/mL) were 148 ± 43 and 196 ± 56, respectively; the plasma concentration was not detected after 6–8 h from the last treatment. The average AUC (ng h/mL) at the 1st and 27th dose of 10 mg/kg azathioprine were 192 ± 31 and 192 ± 30, respectively. These results showed that azathioprine exposure depended on the treatment dose during the dosing period. Furthermore, no larger differences in the TK parameters were observed among the three organizations.

4. Discussion

Although microsampling is applicable to toxicological analysis, concerns exist regarding the effect of serial microsampling from the jugular vein on toxicological parameters. In the present study, we selected azathioprine as the model immunotoxic drug and generated high reliable data including comparison among multiple organizations.

Azathioprine treatment was expected to induce myelosuppression and reduction in thymus size owing to its immunotoxicological property. In this study, erythroid and leukocyte parameters were reduced in rats treated with 10 mg/kg azathioprine. Although the degree of reduction differs depending on the organizations, the group treated with 10 mg/kg azathioprine showed lower average hematocrit and hemoglobin values and lower average count for WBCs, RBCs, neutrophils, lymphocytes, and eosinophils than the non-treatment (0 mg/kg) group (Table 1). The change in erythroid and leukocyte parameters and reduction in thymus size were observed in the azathioprine-treatment groups at each organization. The results of the present study are compatible with those reported by Tochitani et al. [6]. Their report showed the effect of microsampling from rats treated with 12 and 24 mg/kg azathioprine for 14 days in a single organization; this group showed lower red and white blood cell counts than the vehicle group. Furthermore, they showed that, in groups not subjected to microsampling, the hematocrit and hemoglobin values and RBC count in the azathioprine treatment group were significantly lower than that in the vehicle group; however, these significant changes were apparently masked by microsampling. Azathioprine treatment significantly lowered eosinophil and monocyte counts only in the group subjected to microsampling. In our present study, these effects of microsampling were not replicated in any organizations. Generally, the present azathioprine treatment study showed that serial microsampling of 50 μL at 6 points on day 1 and 7 points on day 27 from the jugular vein produced no or minimal influence on clinical signs, body weight, food consumption, hematological parameters, biochemistry parameters, urine parameters, organ weights, and pathological examination in the 28-day assessment of azathioprine as a model immunotoxicological drug. Although almost all parameters showed a similar trend with and without microsampling, changes in two leukocytic, one coagulation, and two biochemical parameters appeared to have been consistently masked by microsampling in two of the three organizations.

Because azathioprine was used as an immunotoxicant, deterioration or masking of immunotoxicological parameters by microsampling were the most important aspects evaluated in this study. We observed that microsampling was associated with apparent masking of reduction in WBC and lymphocyte counts in the group treated with 10 mg/kg azathioprine in organizations B and C; however, this effect was slight in organization A. The reason for this apparent masking is currently unknown. This phenomenon might be related to slight inflammation by frequent puncture for serial microsampling. Thus, this effect should be considered when using microsampling in rat toxicity studies on potentially immunotoxicological drugs. We additionally observed an apparent masking of azathioprine-induced reduction in PT and increase in total cholesterol and Ca levels. The apparent disappearance of PT reduction by microsampling could be caused by enhanced activation of the coagulation system through loss of circulating blood; however, this phenomenon was not observed in our previous study on microsampling without drug administration [3]. Additionally, increasing effects on total cholesterol and Ca values were not detected systemically in the previous study. Thus, the current findings may be related to both azathioprine treatment and microsampling. Further studies using other immunotoxicants are necessary for replication.

Azathioprine treatment was known to reduce the thymus weight in rats [7], [8], [9]. We observed that azathioprine-induced systematic decreases in thymus weights and decreased the number of lymphoid cells and lymphocytes in the thymus cortex irrespective of microsampling, suggesting no influence on these evaluations.

The azathioprine exposure was analyzed by the measurement of its metabolite 6-MP using the sample volume of 50 μL. The peak concentrations and AUC levels showed dose-dependency, which resulted in dose-dependent azathioprine toxicity. This result supported the premise that microsampling allows the evaluation of the relationship between toxicological observation and TK in each animal.

A limitation of this study was that we did not aim to strictly assess the same parameters in terms of biochemistry (α1-globulin, γ-globulin, phospholipid, GLDH, LDH, γGT, and guanase; Supplementary Table 2), urine (specific gravity and osmolarity, Supplementary Table 3), organ weights (the lungs; brain; pancreas; uterus; ovaries; and the pituitary, submandibular, adrenal glands; Table 2 and Supplementary Table 4), and pathology (the heart, lungs, subcutaneous tissue, cervical node, intestinal mesenteric lymph node, and brain; Supplementary Table 6). These parameters were not evaluated at all organizations, although routine parameters were evaluated in each organization. Therefore, the evaluation of the effect of microsampling on these parameters might be insufficient, considering that this was a multi-organization study.

5. Conclusions

Azathioprine toxicity could be assessed appropriately as an overall profile, even with blood microsampling. However, microsampling may influence azathioprine-induced changes in a few parameters, especially leukocyte parameters; therefore, its usage should be carefully considered. From these results, we concluded that microsampling would be applied for TK evaluation in the main study animals treated with immunotoxicity drugs, but cautions should be paid in the minor influence on several hematological parameters for both sponsors and regulators. To reduce the influence, more reduced sampling volume and/or better microsampling technique might be beneficial and need to be evaluated. Also, further studies using other immunotoxicants are necessary to better understand the influence on microsampling on toxicological profiles.

CRediT authorship contribution statement

Yoichi Tanaka: Investigation, Writing - original draft, Writing - review & editing. Kazuaki Takahashi: Investigation, Writing – review & editing. Norimichi Hattori: Investigation, Writing – review & editing. Hideaki Yokoyama: Investigation, Writing – review & editing. Koki Yamaguchi: Investigation, Writing – review & editing. Yusuke Shibui: Investigation, Writing – review & editing. Sayaka Kawaguchi: Investigation, Writing – review & editing. Taishi Shimazaki: Investigation, Writing – review & editing. Keiko Nakai: Investigation, Writing – review & editing. Hiroyuki Kusuhara: Investigation, Writing – review & editing. Yoshiro Saito: Investigation, Writing – review & editing.

Funding

This work was supported in part by AMED under Grant nos. JP20ak0101073j0004 and JP21ak0101073j0005.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Handling Editor: Dr. L.H. Lash

Appendix A. Supplementary material

Supplementary material

Data Availability

The authors do not have permission to share data.