2022 Volume 47 Issue 5 Pages 193-199
According to ICH S3A Q&A focusing on microsampling, its application should be avoided in main study animals for test drugs that could exacerbate hematological parameters with frequent blood sampling. However, no study has reported the effects of microsampling on toxicity parameters of drugs known to induce hematological toxicity. Therefore, we assessed the toxicological effects of serial microsampling on rats treated with phenacetin as a model drug. In a common 28-day study, 50 µL of microsampling was performed at 6-time points on days 1 to 2 and 7-time points on days 27 to 28 from the jugular vein of Sprague Dawley rats. The study was performed independently by two organizations. The toxicological influence of microsampling was evaluated on body weight, food consumption, hematology, blood clinical chemistry, urine parameters, organ weights, and tissue pathology. Phenacetin treatments induced significant changes of various hematological parameters (including hemoglobin and reticulocytes), some organ weights (including liver and spleen), and some hematology-related pathological parameters in the liver, spleen and bone marrow. Meanwhile, serial microsampling exhibited minimal influence on the assessed parameters, although 20 parameters showed statistical differences mostly at one organization. The current results support the notion that serial 50 μL microsampling from the jugular vein had minimal impacts on overall toxicological profiles even in rats treated with a drug inducing hematological toxicity, but the potential adverse effect on certain parameters could not be fully excluded. Accordingly, this microsampling technique has possibility to be employed even for non-clinical rat toxicity studies using drugs with potentially hematological toxicity.
Rats are commonly used for non-clinical safety studies, and toxicokinetic (TK) evaluations are persistently performed to assess the relationship between toxicological data and systematic exposure to drug candidates based on ICH S3A guidelines (ICH, 1994). In rodent studies, satellite animals are frequently established for sampling of TK analysis, with 100–200 μL of blood collected at each time point. However, collection of these samples can impact toxicological parameters in the main study animals. In November 2017, ICH S3A Q&A focused on microsampling was released (ICH, 2017) to accelerate microsampling for TK evaluation. Typically, microsampling is characterized by the collection of ≤ 50 μL blood at each time point, which is now sufficient for TK evaluation, given the progress in apparatus used for measurements.
The S3A Q&A section “3.1 (Q6) How to evaluate the effect of blood sampling on the toxicity data and wellbeing of the animal in main study group?” describes the points at which blood sampling was performed on the main study animals. It is crucial to consider the effect of blood collection on the physiological conditions of animals. If previous studies have shown that test drug-related changes in hematological parameters could be exacerbated by frequent blood sampling, the use of satellite groups of animals for TK assessment would be warranted. Accordingly, the microsampling technique would be avoided in the main study animals for test drugs that can exacerbate hematological parameters with frequent blood sampling. However, no study has reported the effects of actual microsampling on the toxicity parameters of drugs capable of inducing hematological toxicity.
We have previously shown that serial 50 μL × 6 (day 1 to 2) + 7 (day 27 to 28) point-microsampling from the jugular vein of Sprague Dawley (SD) rats had no or only minimal influences on body weight gain, food consumption, hematological and blood clinical chemistry parameters, and organ weights when performed at four independent organizations in a common 28-day study (Yokoyama et al., 2020). In the present study, using the same sampling points as described in the previous study, we assessed the toxicological effects of serial microsampling on rats treated with phenacetin as a model drug known to induce hematological toxicity (Robson, 1965; Boelsterli et al., 1983) at two independent organizations.
Two organizations (A and B), except the National Institute of Health Sciences, performed the predetermined animal experiments. Female SD rats (Crl:CD; 5-weeks of age) were purchased from Charles River Laboratories Japan, Inc. (Yokohama, Japan). Female rats were selected due to their lower circulatory blood volume compared with male rats, presumably causing them to be more significantly impacted by microsampling. Rats were fed CR-LPF (Oriental Yeast Co., Tokyo, Japan) and tap water ad libitum. The study was initiated at each organization after one week of habituation. This study was approved by the Animal Experiment Committee of each organization.
Study protocolTwo rat groups were established at each organization: without (group I) and with (50 μL, group II) microsampling groups, comprising five females each. Each rat was orally administered phenacetin (suspended in methylcellulose) daily at 0, 300, and 1,000 mg/kg/day. Then, 50 μL of blood was collected from the jugular vein of group II rats without anesthetization at 0.5, 1, 2, 4, and 8 hours (hrs) on day 1 and 24 hr, and at 0, 0.5, 1, 2, 4, and 8 hrs on day 27 and 24 hr, after administration. The study was initiated at Time 0, set in the morning. Blood was collected using 1-mL low-dose syringes with 29G needles (Becton Dickinson Co., Franklin Lakes, NJ, USA) and heparin. All animals were assessed daily for clinical signs, and body weight and food consumption were measured at least once weekly. In addition, urine was collected for 4 or 20 hr after phenacetin administration on day 22, as described in Supplementary Table 3.
Rats were fasted from day 28, and on day 29, rats were anesthetized with isoflurane (Mylan V.V., Tokyo, Japan, or Pfizer Inc., New York, NY, USA) and 4–5 mL of blood was withdrawn from the abdominal vena cava for hematological and blood clinical chemistry assessments. Rats were then sacrificed by exsanguination, and organs were harvested and weighed (including the heart, lung, liver, kidneys, thymus, spleen, and adrenal glands). The liver, kidney, urinary bladder, thymus, spleen, bone marrow, lung, and adrenal gland tissues (at maximum) were fixed using formalin and embedded in paraffin for pathological examination.
The following hematological parameters were assessed: red blood cell (RBC) count, hematocrit, hemoglobin, mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), reticulocyte count, mean corpuscular volume (MCV), red blood cell distribution width (RDW), leukocyte/white blood cell, neutrophils, lymphocytes, monocytes, basophils, eosinophils, platelet counts, fibrinogen, prothrombin time (PT), and activated partial thromboplastin time (APTT) (at maximum). The measured blood clinical chemistry parameters were total protein, albumin, albumin/globulin (A/G) ratio, glucose, total cholesterol, triacylglycerol, alanine transaminase (ALT), aspartate transaminase (AST), glutamate dehydrogenase (GLDH), alkaline phosphatase (ALP), total bilirubin, direct bilirubin, lactate dehydrogenase (LDH), γ-glutamyltranspeptidase (gammaGT), creatine kinase (CK), blood urea nitrogen (BUN), creatinine (Cre), Na, K, Ca, inorganic P, and Cl (at maximum). Urine parameters were assessed for volume, specific gravity, Cre, total protein, albumin, Na, K, and Cl (at maximum).
TK evaluation of phenacetinPhenacetin was extracted and measured according to the method described by Whiterock et al. (2012), with slight modifications. For extracting phenacetin, plasma (5 μL) was mixed with 200 μL of an internal standard (Phenacetin-d5 in acetonitrile). The mixture was centrifuged at 15,000 rpm (organization A) or 10,000 rpm (organization B) for 5 min at 10°C (organization A) or 4°C (organization B), and the supernatant (40 μL) was transferred to another tube and mixed with 160 μL of acetonitrile:water (1:1, v/v). Phenacetin levels were measured using liquid chromatography-tandem mass spectrometry. The liquid chromatography system used was LC-30AD (organization A) or LC-20A (organization B) from Shimadzu (Kyoto, Japan) with a Cadenza CD-C18 (3 μm, 2.0 × 30 mm) column (Imtakt, Kyoto, Japan) maintained at 60°C. Mobile phase A was composed of water:formic acid (1000:1, v/v) and mobile phase B was acetonitrile:formic acid (1000:1, v/v). Separation was achieved using a gradient of 5% (0 min) to 50% (0.7 min) of mobile phase B, and then this concentration was maintained at 1.2 min, at a flow rate of 0.6 mL/min. Thereafter, the column was washed by increasing mobile phase B to 95% at 1.21 min, maintained at 1.7 min, and then re-initialized by decreasing mobile phase B to 5% at 1.71 min through 2.7 min. The interfaced mass spectrometer used was an API 4000 (organization A, AB Sciex, Framingham, MA, USA) or QTRAP 5500 (organization B, AB Sciex), operated in the positive electrospray ionization mode using multiple reaction monitoring.
Statistical analysisAt each organization, statistical difference for each parameter following treatment with 0 and 300 mg/kg or 1,000 mg/kg phenacetin was evaluated by Williams test (for equal variances) or Shirley-William test (for non-equal variances) after evaluation by Bartlett test. If the result of Williams test or Shirley-William test did not reach significance, Dunnett test or Steel test was applied, respectively. Also, statistical comparisons between 0 (group I) and 50 μL (group II) sampling groups were performed using Student’s t-test or Welch’s t-test after equality evaluation by F-test. Correction for multiple comparisons was not applied to determine minor influences of microsampling.
Supplementary Table 1 presents the body weight and food consumption evaluated every week. Body weights gradually increased in both groups I and microsampling group II. Apparent deterioration of reduced body weight gain by microsampling was observed in rats treated with phenacetin 1,000 mg/kg at organization B, however, a similar phenomenon was not found by organization A. At each organization, no statistical difference was observed between groups I and II (with or without microsampling) at every assessment point. For food consumption, significant decreases were observed in phenacetin 1,000 mg/kg groups compared with non-treatment group at both organizations. Regarding the influence of microsampling, group II at organization B showed statistically more consumption at 1 (1,000 mg/kg), 2 (0 and 1,000 mg/kg) and 4 weeks (300 mg/kg) than those of group I, but these differences were small and no statistical significance was observed for all weeks at the other organization. These results suggested that the 50 μL microsampling had no or minimal influence. Thus, the obtained toxicity profile for phenacetin regarding body weight and food consumption could be mostly assessed appropriately, even when blood was collected via microsampling.
Table 1 presents the results of hematological parameters at week 4. As the phenacetin dosage was escalated, RBC counts, hematocrit, hemoglobin and MCHC values decreased, while MCH, reticulocyte and MCV (and RDW at one organization) increased, in dose-dependent manner, with statistical significance when assessed at both organizations, indicating hematological toxicity. By microsampling, the statistical significance by the phenacetin treatments was maintained for RBC counts, MCH, MCHC, reticulocytes, MCV and RDW. Exceptions were hematocrit in phenacetin 1,000 mg/kg at both organizations and hemoglobin in phenacetin 300 mg/kg at organization A. There was no apparent deterioration or masking of phenacetin effects on hematological toxicity by microsampling with same trends in the two organizations. Additionally, lower RBC and higher reticulocyte values were consistently observed in group II compared with group I. This may have been due to blood loss associated with serial microsampling; however, these differences were not consistently significant between groups I and II, probably due to slightness of these effects. Among the 18 tested parameters, MCH, MCV, RDW, and basophil values of group II at 1,000 mg/kg phenacetin significantly differed from those of group I in organization A. In addition, fibrinogen value of group II at 300 mg/kg phenacetin increased compared with that of group I in organization A. On the other hand, RBC, hemoglobin, MCHC, reticulocyte, and MCV values of 0 mg/kg group II rats at organization B significantly differed from those of the corresponding group I. However, these statistical differences between groups I and II were not observed in the corresponding groups assessed at the other organization. These results suggest that the influence of the microsampling treatment was minimal and not reproducible between organizations (not in more than one organization per condition) and the hematological parameter profile of phenacetin could be mostly assessed appropriately even in the rats that had undergone blood microsampling.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/47_193-t1.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
We next analyzed the blood clinical chemistry (Supplementary Table 2). Influence by phenacetin treatment was significantly observed in several parameters including ALT, total bilirubin, direct bilirubin and total cholesterol at both organizations, indicating mild liver toxicity. Also, slight but statistically significant changes induced by phenacetin were seen on other parameters including CK at organization B, but not at organization A. For influence of microsampling, sporadic but significant differences for three parameters, Na, Ca and Cl, between groups I and II (observed only at organization B per parameter). The Cl values for all rats in group II were slightly higher than those of group I at organization B, but no dose-dependent trend with phenacetin dosage was noted, and no significant changes was shown at organization A. We regarded that the phenacetin toxicity profile on blood clinical chemistry could be mostly assessed appropriately even in rats that had undergone blood microsampling. However, it is noteworthy that increase in ALT induced by phenacetin 1,000 mg/kg treatment (3.6–4.2 fold from 0 mg/kg rats of group I) was apparently undermined (i.e., partially masked) by microsampling (2.4–3.0 fold) in both organizations, although there were no statistical significance between groups with or without microsampling (Group I vs. group II).
Next, we performed urine analysis (Supplementary Table 3). Phenacetin treatment significantly altered urine volume (at both organizations) as well as Cre and albumin values (at organization A), suggesting mild kidney injury. Statistically significant increases in Na and Cl values were also observed in the rats administered 1,000 mg/kg phenacetin only at organization B. Regarding influence of microsampling, we observed that four parameters, urine volume, Na, K and Cl, differed significantly between groups I and II for at least one phenacetin dosage at either organization. Among these, urine volume of microsampled rats (group II) administered 1,000 mg/kg phenacetin at organization B was significantly increased when compared with no microsampled rats (group I). This deterioration might be caused by microsampling, but considering that significantly decreased urine volume was observed at group II treated with 300 mg/kg phenacetin at organization B, and similar trend was not observed at organization A, we regarded this phenomenon as independent of microsampling. Moreover, the Cl values of group II at 300 mg/kg phenacetin significantly differed from those in group I at both organizations; however, the trends tended to vary. That is, in group II, the value was increased in the assessment by organization A, but decreased at organization B, when compared with those in group I. These results suggested that the observed differences were random, and the phenacetin toxicity profile on urine parameters could be mostly assessed appropriately even in the rats that had undergone blood microsampling.
We next examined organ weight (Supplementary Table 4). Phenacetin treatment dose-dependently increased liver and spleen weights with significance at both organizations. In addition, thymus weight was significantly decreased in 1,000 mg/kg-treated rats at both organizations. These phenacetin effects were observed also in data on organ weights per body weights (Supplementary Table 5). These phenacetin toxicities could be mostly assessed appropriately as a profile, even when blood was taken by microsampling. In 1,000 mg/kg phenacetin-treated rats, the liver and thymus weights were significantly increased in group II rats when compared with those of group I rats, as determined at organization A; however, these values inversely decreased when determined at organization B, with no statistical significance. Of note, the statistical significance of increased liver weights following phenacetin (300 and 1,000 mg/kg) treatment in organization B disappeared following microsampling; however, the overall trend differed between the two organizations. Supplementary Table 5 presents data on organ weights per body weights, and increased values for the thymus (1,000 mg/kg dose at organization A) and decreased values for liver and kidney (300 mg/kg dose in organization B) were statistically significant by microsampling; however, these values demonstrated an inverse tendency or minimal change with no significance in the other organization. Note that the disappearance of phenacetin effect on liver weight by microsampling (Supplementary Table 4) was not detected in this assessment by organ weights per body weights (Supplementary Table 5) at organization B.
The final comparison was a pathological examination (Table 2). As expected from its hematological toxicity, phenacetin treatment caused pigment deposition (brown, Kupffer cell) in the liver, congestion, extramedullary hematopoiesis, increased pigment deposition (brown) in the spleen, and increased hematopoietic cells in the bone marrow at both organizations. Moreover, atrophy in thymus was observed at organization A. All the observed pathological toxicities were minimal or mild, and their grades were not influenced by microsampling; i.e., there was no difference in the pathological grades in the tissues related to anemia (extrinsic hematopoiesis and brown pigmentation) with or without microsampling (data not shown). Thus, microsampling induced no obvious change in any of the evaluated pathological parameters. Hence, these phenacetin effects could be mostly assessed appropriately, even in the rats that had undergone blood microsampling.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/47_193-t2.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
In addition, we evaluated the influence of microsampling on the general performance of animals. Following 1,000 mg/kg phenacetin administration, decreased locomotor activity was detected in 2 and 0 rats in group I and 3 and 2 rats in group II among five rats examined in each group at organizations A (on day1) and B (on day 1 and day 22), respectively. Therefore, we cannot exclude the possible impact of microsampling in the present evaluation, although this was not reflected in the other evaluated parameters. Furthermore, we examined TK of phenacetin in 3 out of 5 rats in each group at both organizations and confirmed its dose-dependent exposure (Supplementary Table 6).
The current results suggest that the phenacetin toxicity could be mostly assessed appropriately as overall profiles even with blood microsampling. Serial 50 μL × 6 (day 1 to 2) + 7 (day 27 to 28) point-microsampling from the jugular vein of SD rats had mostly minimal influence on body weight gain, food consumption, hematological parameters, blood clinical chemistry, urine parameters, organ weights, and tissue pathology in a 28-day assessment of rats treated with the hematotoxic drug phenacetin. Blood volumes collected on days 1–2 and days 27–28 corresponded to approximately 2.8% and 2.3~2.7% of the total blood volume, respectively, when blood volume per kg body weight for rats was set as 64 mL/kg (Diehl et al., 2001). We detected statistically significant differences between groups I and II in the values for 20 parameters, mostly either at one of the organizations but not at the other organization, or different trends at the two organizations, suggesting that the observed differences were random and not systematically induced due to microsampling. Nevertheless, minimal effects were observed following microsampling on a few hematological parameters. A precise evaluation of the results highlighted that RBC counts were slightly decreased, and reticulocyte values were slightly increased in the microsampling group (group II) at both dosages and organizations (Table 1). However, these phenomena were also observed in rats not administered phenacetin. In addition, we have previously reported a similar influence in microsampled female rats at four organizations (Yokoyama et al., 2020). Thus, these effects on RBC and reticulocytes were possibly attributed to blood loss caused by microsampling, irrespective of phenacetin administration, and seemed to be minimal for phenacetin toxicity evaluation. However, we could detect apparent deterioration or masking of phenacetin toxicity by microsampling in certain sporadic parameters (including body weight, and liver weight), but primarily at one organization. The exception was that increases in ALT values followed by phenacetin administration were apparently and partially masked by microsampling at both organizations, although, statistical significance was maintained. Therefore, we have concluded that phenacetin overall toxicity profile could be appropriately evaluated even with microsampling, but we could not fully exclude the possibility that microsampling may have impacts on toxicity deterioration or masking by the test substances. Further studies using other types of test drugs are warranted to better understand the microsampling effects with toxic substances.
One limitation of the present study was that we did not plan to strictly evaluate the same parameters in terms of hematology (RDW at organization B, Table 1), urine (Cre, total protein and albumin in organization B, Supplementary Table 3), and pathology (urinary bladder, lung, and adrenal gland at organization A, and atrophy in thymus at organization B, Table 2). Notably, evaluating the effects of microsampling on these parameters would be insufficient as a multi-organization study. Another limitation is that microsampling was performed by different researchers in the different laboratories, and thus microsampling technique of the two laboratories were not at the same level as each other, although all the researchers had plenty of experience in microsampling. Thus, the results of this study may include bias regarding difference in the microsampling technique.
The current results could accelerate the use of microsampling techniques in rat toxicity studies for drug candidates with no established toxicity profiles in pharmaceutical companies/contract research organizations.
This work was supported in part by AMED under Grant Numbers JP19ak0101073j0003 and JP20ak0101073j0004.
Conflict of interestHirohiko Ohtsuka, Harumi Kitaura, Hitoshi Kandori, Kenta Danbayashi, Tomoaki Higuchi, and Fumihiro Jinno are employees of Axcelead Drug Discovery Partners Inc. Kanae Mori and Shin-ichiro Nitta are employees of the LSI Medience Co. Kazuaki Takahashi, Keiko Nakai, and Atsushi Iwai are employees of the LSIM Safety Institute Corporation. No other conflicts of interest are declared in this work.
