Toxicologic Pathology, XX: 1-15, 2014 Copyright # 2014 by The Author(s) ISSN: 0192-6233 print / 1533-1601 online DOI: 10.1177/0192623314559103

Toxicity of Hydroxyurea in Rats and Dogs DANIEL MORTON1, LORI REED1, WENHU HUANG2, JOHN M. MARCEK3, ROBERT AUSTIN-LAFRANCE3, CARRIE A. NORTHCOTT3, SCOTT H. SCHELLING1, BRADLEY E. ENERSON3, AND LINDSAY TOMLINSON1 1

Pfizer Inc., Andover, Massachusetts, USA 2 Pfizer Inc., La Jolla, California, USA 3 Pfizer Inc., Groton, Connecticut, USA ABSTRACT

The toxicity of hydroxyurea, a treatment for specific neoplasms, sickle-cell disease, polycythemia, and thrombocytosis that kills cells in mitosis, was assessed in repeat-dose, oral gavage studies in rats and dogs and a cardiovascular study in telemetered dogs. Hydroxyurea produced hematopoietic, lymphoid, cardiovascular, and gastrointestinal toxicity with steep dose response curves. In rats dosed for 10 days, 50 mg/kg/day was tolerated; 500 mg/kg/day produced decreased body weight gain; decreased circulating leukocytes, erythrocytes, and platelets; decreased cellularity of thymus, lymph nodes, and bone marrow; and epithelial degeneration and/or dysplasia of the stomach and small intestine; 1,500 mg/kg/day resulted in deaths on day 5. In dogs, a single dose at 250 mg/kg caused prostration leading to unscheduled euthanasia. Dogs administered 50 mg/kg/day for 1 month had decreased circulating leukocytes, erythrocytes, and platelets; increased bone marrow cellularity with decreased maturing granulocytes; increased creatinine kinase activity; and increased iron pigment in bone marrow and hepatic sinusoidal cells. In telemetered dogs, doses 15 mg/kg decreased systolic blood pressure (BP); 50 mg/kg increased diastolic BP, heart rate, and change in blood pressure over time (þdP/dt), and decreased QT and PR intervals and maximum left ventricular systolic and end diastolic pressures with measures returning to control levels within 24 hr. Keywords:

bone marrow; blood; dog; heart; hydroxyurea; rat.

in patients with sickle-cell disease. Mechanisms of action that may benefit sickle-cell disease patients include increasing the production of fetal hemoglobin in erythrocytes, decreasing circulating neutrophils, increasing water content of erythrocytes, enhancing the deformability of erythrocytes, altering adhesion of erythrocytes to endothelium, promoting vasodilation through production of nitric oxide and similar molecules, and increasing adenosine deaminase activity in circulating monocytes leading to lower adenosine levels (Food and Drug Administration [FDA] 2012; Silva-Pinto et al. 2014; Vankayala et al. 2012). Hydroxyurea is also a metal chelator, an inhibitor of catalase, and a substrate for catalase (Italia, Colah, and Ghosh 2013; Juul et al. 2010). Although generally tolerated, hydroxyurea has a low margin of safety with leukopenia and myelosuppression reported as the most common dose-limiting adverse findings in human patients (Charache et al. 1995; FDA 2012). Anemia is the most commonly observed toxicity in dogs (Rassnick et al. 2010). In addition to bone marrow toxicity, growth retardation, spermatogenic arrest, and hemosiderosis of liver and spleen have been described in dog toxicity studies (FDA 2012). In rats administered high doses, testicular atrophy, decreased sperm production, decreased fertility in males, hepatic damage with fatty metamorphosis, and acute necrosis of lymphoid organs and gastrointestinal epithelium have been described (FDA 2012; Frederick et al. 1967). Hydroxyurea is labeled as a genotoxic carcinogen (FDA 2012). Cardiovascular effects and slight methemoglobinemia were reported in some species ‘‘at doses exceeding clinical levels’’ (FDA 2012). Cutaneous and nail lesions have been observed in human and canine patients

INTRODUCTION Hydroxyurea has been used for many years to treat certain neoplasms, polycythemia, and thrombocytosis in humans and dogs. In the United States, hydroxyurea is approved for the treatment of melanoma, chronic myelogenous leukemia, ovarian cancer, and squamous neoplasia of the head and neck, but it has also been used to treat renal cell carcinoma, transitional carcinoma of the urinary bladder, prostatic carcinoma, and carcinoma of the uterine cervix (Liebelt et al. 2007). Hydroxyurea kills dividing cells in S phase of cell division, and this contributes to antineoplastic activity and reduced production of hematopoietic cells (Philips et al. 1967). Since 1998, hydroxyurea has been the only drug approved for reducing painful vaso-occlusive crises, blood transfusions, and hospitalization The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. The author(s) received no financial support for the research, authorship, and/or publication of this article. Address correspondence to: Daniel Morton, Pfizer Inc., Drug Safety Research and Development, 1 Burtt Road, F1137, Andover, MA 01810, USA; e-mail: [email protected]. Abbreviations: AUC, area under the plasma drug concentration curve; BP, blood pressure; Cmax, maximal plasma drug concentration; cGMP, cyclic guanosine monophosphate; DNA, deoxyribonucleic acid; dP/dt, change in blood pressure over time; ECG, electrocardiogram; FDA, Food and Drug Administration; H&E, hematoxylin and eosin; HPD, hours postdose; IV, intravenous; LC/MS/MS, liquid chromatography/mass spectroscopy/mass spectroscopy; LOAEL, lowest observed adverse effect level; LV, left ventricular; NOAEL, no observed adverse effect level; QTc, corrected QT interval; SD, standard deviation; t1/2, half-life; Tmax, time of maximal plasma drug concentration. 1

Downloaded from tpx.sagepub.com at UNIVERSITE DE MONTREAL on September 6, 2015

2

MORTON ET AL.

(Marconato, Bonfanti, and Fileccia 2007; O’Branski et al. 2001; Vassallo et al. 2001). In developing novel therapies for sickle-cell disease that either augment or replace hydroxyurea, knowledge of the toxicity of hydroxyurea and maximum tolerated doses in species normally used for nonclinical toxicity studies is useful for predicting and evaluating toxicity of combination therapies and demonstrating superior safety of a novel therapy compared to hydroxyurea. Very few detailed reports of the adverse effects of hydroxyurea in standard nonclinical toxicity studies in laboratory species have been published. Maximum tolerated doses following administration of multiple doses are not clearly described (FDA 2010, 2012; Liebelt et al. 2007). In the Food and Drug Administration (FDA 2012) label for hydroxyurea (Droxia1, Bristol-Myers Squibb, Princeton, NJ), hydroxyurea administered to rats at 5,780 mg/kg was lethal following a single dose, and repeated doses up to 2,520 mg/kg/day in rats for 40 days produced toxicity with no mention of deaths. In dogs, dose levels up to 1,260 mg/kg/week administered 3 or 7 days weekly for 12 weeks produced ‘‘growth retardation, slightly increased blood glucose values, and hemosiderosis of the liver or spleen’’ and spermatogenic arrest, while deaths were not mentioned (FDA 2012). Published reports were insufficient to clearly identify optimal dose levels for use in future nonclinical toxicity studies. This article describes the toxicity of hydroxyurea in rats and dogs in studies lasting up to 1 month. METHOD Animals and Husbandry Male and female Sprague-Dawley rats (Crl:CD1 [SD], Strain Code 001, Charles River Laboratories, Kingston, NY) were 10 to 12 weeks old at study initiation. Rats were housed in solidbottom polycarbonate cages with paper bedding (ALPHA dri1, Shepherd Specialty Papers, Richland, MI). Male and female Beagle dogs at least 8 months of age and weighing approximately 5 to 11.5 kg at dosing initiation were supplied by Marshall Bioresources (North Rose, NY) and housed individually in 4 ft  6 ft runs. All studies were performed in facilities accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care International in compliance with federal and local regulations. All procedures were approved by the local animal care and use committees. Exploratory Oral Toxicity Study in Rats Male and female Sprague-Dawley rats (5/sex/group) were administered hydroxyurea dissolved in purified water at 50, 500, or 1,500 mg/kg/day by gavage for up to 10 days. The high-dose level of 1,500 mg/kg/day was chosen because administration of hydroxyurea to rats at doses of 2,520 mg/kg/day for 40 days was reported to cause toxicity, but no deaths were reported (FDA 2012). The dose level of 500 mg/kg/day was expected to produce adverse effects, and 50 mg/kg/day was expected to produce minimal or no toxicity. Control animals (5/sex) received only purified water.

TOXICOLOGIC PATHOLOGY

The dose volume for all animals was 10 ml/kg/day. Body weight was assessed twice weekly, food consumption was measured weekly, and clinical signs were monitored twice daily. At a dosage of 1,500 mg/kg/day, all rats died or were electively euthanized in moribund condition on day 5. These animals were necropsied, but clinical pathology, organ weight, and microscopic evaluations were not performed. All remaining rats at 50 and 500 mg/kg/day survived until scheduled euthanasia and necropsy on day 10. Hematology samples (collected with potassium [K2]EDTA) and clinical chemistry samples (serum) were collected from the vena cava on day 10 and analyzed on the Siemens Advia 2120 and 1800 (Siemens Healthcare Diagnostics Inc., Tarrytown, NY), respectively. Plasma hydroxyurea concentrations were measured on days 1 and 9 at approximately 1-, 2-, 4-, and 24-hours postdose (HPD). Rats were euthanized by exsanguination while deeply anesthetized with isoflurane. At necropsy, the liver, spleen, and thymus of each animal in control, 50-, and 500-mg/kg/day groups were weighed. Selected tissues were examined microscopically from control animals and animals administered 500 mg/kg/day. Target organs identified at 500 mg/kg/day were examined in rats administered 50 mg/kg/day. 2-Week Oral Range Finding Study in Dogs Male and female Beagle dogs (1/sex/group) were administered hydroxyurea dissolved in purified water at 50, 250, or 1,000 mg/kg/day. The FDA label for hydroxyurea (Droxia1, Bristol-Myers Squibb, Princeton, NJ) states that 1,260 mg/kg/ week given 3 days a week (420 mg/kg/dose) or every day (180 mg/kg/dose) for 12 weeks produced toxicity, but does not mention deaths (FDA 2012). The maximum tolerated dose could not be found in published literature. Doses for this small range finding study were selected with the expectation that 1,000 mg/kg/day likely would exceed the maximum tolerated dose, 250 mg/kg/day would produce toxicity, and 50 mg/kg/ day (within the therapeutic dose range in dogs) would be tolerated. A vehicle control group of animals (1/sex) received purified water only. The dose volume for all animals was 10 ml/kg. All dogs administered 250 or 1,000 mg/kg/day were euthanized in moribund condition within hours after receiving the first dose; blood oxygen saturation was measured with a tongue sensor (SurgiVet1 V3395 TPR Monitor, Smiths Medical PM, Inc., Norwell, MA) in animals administered 250 mg/kg based on clinical signs of prostration and blue gums suggesting low blood oxygenation; hematology (with K3EDTA) and serum clinical chemistry assessments and necropsies were performed for dogs administered 250 mg/kg; toxicokinetic and microscopic evaluations were not conducted. In all animals in the control group and animals administered 50 mg/kg/day, clinical signs were recorded 2 times a day, body weights were recorded weekly, and qualitative assessment of food intake was recorded daily. Hematology and clinical chemistry samples collected from the cephalic vein prior to initiation of dosing and on day 15 were analyzed on Siemens Advia 120 and 1200 analyzers (Siemens Healthcare

Downloaded from tpx.sagepub.com at UNIVERSITE DE MONTREAL on September 6, 2015

Vol. XX, No. X, 2014

HYDROXYUREA TOXICITY

Diagnostics Inc., Tarrytown, NY), respectively. Plasma hydroxyurea concentrations were measured on days 1 and 14 at approximately 1, 2, 4, 7, and 24 HPD. At the end of the dosing phase on day 15, control dogs and dogs administered 50 mg/kg/ day were euthanized with phenobarbital followed by exsanguination and necropsied. Selected tissues were collected, processed, and examined microscopically from control dogs and dogs administered 50 mg/kg/day. 1-Month Oral Toxicity Study in Dogs Male and female Beagle dogs (3/sex) were administered hydroxyurea dissolved in purified water at 50 mg/kg/day by oral gavage once daily for 28 (males) or 29 (females) days. A vehicle control group of the same size was administered only purified water. The dose volume for both groups was 2 ml/kg/dose. During the dosing phase, clinical signs were recorded 3 times a day, body weights were recorded weekly, and qualitative assessment of food intake was recorded daily. Ophthalmic examinations were performed before administration of the first dose and near the end of the study. Prior to the initiation of dosing and near the end of the study, hematology samples collected with K2EDTA were analyzed using a Siemens ADVIA 1800; coagulation samples collected with sodium citrate were analyzed on STA Compact (Diagnostica Stago Inc, Parsippany, NJ); clinical chemistry samples (serum) were analyzed using a Siemens ADVIA 2120; and urinalysis samples collected via cystocentesis were analyzed on a Siemens Clinitek ATLAS (Siemens Healthcare Diagnostics Inc., Tarrytown, NY). All blood was collected from the jugular vein. Morphology of blood cells was evaluated in blood smears from all animals at the end of the study. Heart rate and lead 2 electrocardiograms (ECGs) using superficial electrodes were monitored in conscious restrained dogs prior to the initiation of dosing and on day 23 prior to and 1 hr after dosing. Blood samples were collected from all animals for measurement of hydroxyurea plasma concentrations and determination of toxicokinetic parameters on days 1 and 27 at approximately 0.5, 1, 2, 7, and 24 hr after dosing. Urine was collected for analysis by cystocentesis at necropsy. At the end of the study, animals were euthanized with pentobarbital followed by exsanguination and a necropsy was conducted. A comprehensive list of tissues was examined microscopically. Histologic Processing Eyes from the exploratory study in rats and the 2-week study in dogs were fixed in Davidson’s fixative. Eyes and optic nerves from the 1-month dog study were fixed in 3% glutaraldehyde. Testes and epididymides from dogs and rats in all studies were fixed in modified Davidson’s fixative. All other tissues from all general toxicity studies were fixed in 10% neutral buffered formalin. Sternum and sternal bone marrow from rats and dogs in all studies and distal tibia from the 1-monthdog study were decalcified prior to trimming and processing. All tissues to be examined by light microscopy were processed

3

routinely, embedded in paraffin, sectioned at approximately 5 mm, and stained with hematoxylin and eosin (H&E). Cardiovascular Safety Pharmacology Telemetry Study in Dogs Male Beagle dogs (N ¼ 4) implanted with a telemetry device (TL11M3-D70-PCTP; Data Sciences International, Minneapolis, MN) were used to evaluate the effects of oral administration of hydroxyurea (0, 1, 15, and 50 mg/kg; 2 ml/kg dissolved in purified water; crossover design) on arterial blood pressure (BP), left ventricular (LV) pressure, ECG (lead 2 configuration), and activity measures. Dogs were surgically prepared prior to the study as follows: A pressure catheter was placed in the femoral artery and advanced to the descending aorta; a second pressure catheter was passed through the femoral artery into the left ventricle through the apex; ECG leads were secured in a lead 2 configuration with the negative electrode attached to the pericardium near the right atrium, and the positive electrode secured to the epicardium at the apex of the left ventricle; the body of the transmitter was implanted into an intramuscular pocket on the left flank of the dog. The high dose was chosen because it is a tolerated, minimally toxic dose based on the hematologic findings at 50 mg/kg/day and acute lethality at 250 in the range finding studies described in this article, and published literature describing very mild hematologic effects at 15 mg/kg/day for 2 weeks and more severe toxicity at doses between 140 and 1,260 mg/kg/week (FDA 2012). Telemetered signals were acquired continuously beginning *1 hr predose through *24 HPD. Measured parameters included systolic, diastolic, and mean BP; LV maximum systolic pressure (an index of hypotension or hypertension), LV end diastolic pressure (an index of LV function), and maximum rise in LV pressure (LV change in blood pressure over time [þdP/dt] maximum, an index of cardiac contractility); heart rate; RR, PR, QRS, QT, and corrected QT (QTc) intervals; and activity. Telemetered data were further reduced to 4 postdose periods (0.75–3.75 HPD, 5.00–9.00 HPD, 9.25– 18.00 HPD, and 18.25–24.00 HPD) for statistical analyses (analysis of variance; p < .05). Plasma samples for determination of hydroxyurea concentration were obtained predose and *4 HPD during each of the 4 treatments in the crossover dosing phase. Following completion of the crossover phase, all animals received a single oral dose of hydroxyurea (50 mg/kg) to obtain a toxicokinetic profile. Plasma samples were obtained predose and *0.5, 1, 2, 4, and 24 HPD during the toxicokinetic phase. Toxicokinetic Assessments For all studies, blood was collected into tubes containing either K3EDTA (rat and dog exploratory studies) or K2EDTA (1-month dog and dog cardiovascular studies) over a 24-hr period for each plasma concentration curve. Plasma was separated by centrifugation and stored at 70 C until hydroxyurea concentrations were measured using a validated

Downloaded from tpx.sagepub.com at UNIVERSITE DE MONTREAL on September 6, 2015

4

MORTON ET AL.

TOXICOLOGIC PATHOLOGY

TABLE 1.—Mean (+ standard deviation) clinical pathology findings in rats caused by hydroxyurea. Males

Females

N ¼ 5/group

N ¼ 5/group

Dose (mg/kg/day)

Dose (mg/kg/day)

0 3

White blood cells (10 /ml) Neutrophils (103/ml) Lymphocytes (103/ml) Monocytes (103/ml) Eosinophils (103/ml) Basophils (103/ml) Red blood cells (106/ml) Hemoglobin (mg/dl) Hematocrit (%) Mean corpuscular hemoglobin concentration (g/dl) Reticulocytes (103/ml) Red cell distribution width (%) Platelets (103/ml) Alkaline phosphatase (U/L) Cholesterol (mg/dl) Total protein (g/dl) Albumin (g/dl) Globulin (g/dl) Potassium (mmol/L) Chloride (mmol/L)

10.6 + 1.3 + 8.6 + 0.45 + 0.086 + 0.02 + 7.1 + 13.6 + 40.8 + 33.2 + 270 + — 1,257 + 239 + 62 + 5.9 + 3.7 + 2.2 + 4.9 + 101 +

1.9 0.5 1.6 0.16 0.02 0.01 0.2 0.3 0.7 0.5 29 111 37 7 0.2 0.2 0.1 0.5 1

50

500

7.2 + 1.3* 0.9 + 0.2* 5.9 + 1.3* 0.27 + 0.05* 0.068 + 0.02 0.010 + 0.01* —a — — — — — — — — — — — — —

3.9 + 1.3* 0.1 + 0.1* 3.8 + 1.2* 0.01 + 0.01* 0.04 + 0.01* 0.00 + 0.00* 5.5 + 0.8* 11.0 + 1.4* 31.3 + 3.8* 35.1 + 0.8* 0 + 0* — 193 + 138* 123 + 33* 82 + 6 4.9 + 0.9* 3.1 + 0.5* 1.8 + 0.3* 4.5 + 0.3 106 + 3*

0

50

+ 4.5 + 0.4 + 4.0 + 0.05 + 0.02 + 0.02 + 0.4 + 0.7 + 2.0 — 255 + 31 11.1 + 0.3 1297 + 118 — — — — — 5.1 + 0.3 —

6.8 + 2.4 0.7 + 0.6 5.6 + 1.9 — — 0.010 + 0.01 — — — — — — — — — — — — — —

8.3 1.0 7.0 0.21 0.080 0.014 6.8 13.0 38.0

500 3.5 0.2 3.2 0.02 0.020 0.004 5.5 10.5 30.2 1 10.5 441

4.5

+ 0.8* + 0.1* + 0.7* + 0.01* + 0.04 + 0.01 + 0.2* + 0.3* + 1.2* — + 2* + 0.1* + 59* — — — — — + 0.2* —

Note. No hydroxyurea-related findings were observed for mean corpuscular volume, mean corpuscular hemoglobin, mean platelet volume, alanine aminotransferase, aspartate aminotransferase, g-glutamyltransferase, glutamate dehydrogenase, albumin/globulin ratio, total bilirubin, blood urea nitrogen, creatinine, calcium, glucose, phosphorus, and sodium. a Denotes no test article–related finding. *Statistically significant at p < .05.

liquid chromatography/mass spectroscopy/mass spectroscopy (LC/MS/MS) method. The maximal plasma drug concentration (Cmax) and area under the drug concentration curve (AUC0–24) were determined for each surviving animal in the exploratory studies and 1-month toxicity study in dogs. RESULTS Toxicity in Rats Test article–related clinical signs observed for males and females at 1,500 mg/kg/day were first noted on day 4 or day 5 and included decreased activity, thin appearance, decreased skin turgor, decreased feces, watery feces, and hunched posture. Aggression, chromodacryorrhea, excessive lacrimation, and ptosis were seen in males only at 1,500 mg/kg/day. By day 5, mean body weights were 22% lower in males and 18% lower in females compared to day 1 mean body weights. Body weight loss correlated with male and female group mean food consumption values that were 57% and 72% lower than control means, respectively, on day 5. Two males were not dosed on day 5 because of the severity of clinical signs and body weight loss, and those 2 rats died on that day. Due to the severity of the clinical observations, all surviving male and female animals administered hydroxyurea at a dose of 1,500 mg/kg/day were euthanized on day 5. At necropsy, 2 of the 5 females at 1,500 mg/kg/day had dark lungs, and 1

of those 2 females also had a dark liver. Gelatinous content or watery content was observed in the small and large intestines of males and females. All males and females at 1,500 mg/kg/day had small spleens and thymuses. Clinical pathology and organ weight data were not collected, and microscopic evaluations were not performed for animals administered 1,500 mg/kg/day due to moribundity and early euthanasia. The cause of deaths and severe clinical signs at 1,500 mg/kg/day were not determined. There were no test article–related deaths at 50 or 500 mg/kg/ day. Test article–related clinical observations observed beginning on day 7 or day 9 in males at 500 mg/kg/day included decreased skin turgor, decreased or discolored (pale) feces, and rough hair coat. In addition, group mean body weights in males and females at 500 mg/kg/day were 14% and 4% lower than the control means, respectively. At 500 mg/kg/day between days 1 and 10, males lost approximately 3.3% of body weight on average and mean body weight gain in females was 0.9%, compared to control male and female body weight gains of 12.6% and 5.9%, respectively. These body weight effects correlated with lower group mean food consumption values. There were no test article–related clinical observations, body weight changes, or effects on food consumption for males or females at 50 mg/kg/day. Clinical pathologic findings in rats attributed to administration of hydroxyurea are presented in Table 1. Test article–

Downloaded from tpx.sagepub.com at UNIVERSITE DE MONTREAL on September 6, 2015

Vol. XX, No. X, 2014

HYDROXYUREA TOXICITY

related and dose-related lower values for circulating white blood cells, neutrophils, lymphocytes, monocytes, eosinophils, and basophils occurred at 50 mg/kg/day on day 10. At 500 mg/kg/day, there were lower values for red blood cells, hemoglobin, hematocrit, red cell distribution width, reticulocytes, and platelets on day 10 of the dosing period. At 500 mg/kg/day, neutrophils, monocytes, and reticulocytes were severely depleted, while mean platelets were 0.15 to 0.34 control means. Red blood cells, hemoglobin, hematocrit, and lymphocytes were less severely affected, and these smaller effects can likely be explained by the longer life spans of circulating erythrocytes and lymphocytes compared to granulocytes and platelets. The lower values for circulating white blood cells, neutrophils, lymphocytes, monocytes, eosinophils, basophils, erythrocytes, reticulocytes, and platelets were consistent with pharmacologically mediated decreased cellularity of the bone marrow and lymphoid tissues observed microscopically at 500 mg/kg/day (Lerner et al. 1966). In males at 500 mg/kg/day, mean corpuscular hemoglobin concentration was higher than the control mean, indicating an increase in free hemoglobin likely associated with intravascular hemolysis. Blood smears were not evaluated, and there was no obvious hemolysis in serum samples. In animals administered 500 mg/ kg/day, lower values for total protein, albumin, and globulin in males were attributed to the gastrointestinal lesions (described subsequently), whereas lower potassium in males and females and lower alkaline phosphatase and higher values for cholesterol and chloride in males were secondary to decreased food consumption and poor clinical condition. Test article–related, absolute, and relative (to body weight) thymic weight changes were present in both sexes at 500 mg/ kg/day. Mean absolute thymus weights (0.20–0.25  control mean) and thymus-to-body weight ratios (0.18–0.21  control mean) correlated with small thymuses at necropsy and microscopic evidence of decreased lymphoid cellularity. No test article–related macroscopic or microscopic findings were observed at 50 mg/kg/day. Primary test article–related microscopic findings occurred in the thymus, mesenteric lymph node, bone marrow, liver, stomach, duodenum, jejunum, and ileum at 500 mg/kg/day (Table 2). Secondary test article–related microscopic findings at 500 mg/kg/day were present in the mesentery and lung. In the thymus, there was marked to severe, diffusely decreased lymphoid cellularity in all rats at 500 mg/kg/day (Figure 1). Decreased cellularity was most striking in the cortex, but also affected the medulla and was associated with decreased thymus weights and a macroscopic diagnosis of small thymus. In the mesenteric lymph nodes of 3 males, minimal to moderate, diffusely decreased lymphoid cellularity was evident. There was diffuse moderate (females) to marked (males) decreased cellularity of all bone marrow precursor cell types in all rats at 500 mg/kg/day. In the liver, there was a correlative lack of extramedullary hematopoiesis in males and females. In the stomachs of 3 males, there was mild focal to diffuse dysplasia of the nonglandular mucosal epithelium associated with variably sized keratinocytes, keratinocyte necrosis, formation of bullae, subcorneal accumulation of

5

TABLE 2.—Microscopic findings in rats attributed to administration of hydroxyurea at 500 mg/kg/day. No test article–related microscopic findings were observed at 50 mg/kg/day.

Microscopic observation Stomach Dysplasia Mild Edema Moderate Duodenuma Dysplasia Mild Crypt degeneration Minimal Mild Ileum Dysplasia Mild Crypt degeneration Minimal Mild Bone marrow Decreased cellularity, all cell lines Moderate Marked Thymus Decreased cellularity, lymphoid Marked Severe Mesenteric lymph node Decreased cellularity, lymphoid Minimal Moderate Mesentery Atrophy, adipose tissue Minimal Mild Moderate Liver Lack of extramedullary hematopoiesis Lung Hemorrhage, multifocal Minimal Mild

Males N ¼ 5/group

Females N ¼ 5/group

3

—b

2

—

4

—

1 2

— —

2

—

2 1

1 —

— 5

5 —

— 5

2 3

1 2

— —

— 1 2

2 1 1

5

5

1 1

— —

a Dysplasia was also observed in the jejunum examined in only 1 male with macroscopic observations. b Denotes no test article–related finding.

proteinaceous fluid, and/or erosion. In 1 of the males, there was also focal to diffuse moderate gastric submucosal edema and hemorrhage. These lesions correlated with raised gastric mucosal lesions evident grossly in 2 males. In the duodenum, jejunum, and ileum, including 1 male with a macroscopic finding of dark discoloration of the duodenum with fluid content in the small and large intestine, there was a combination of minimal to mild multifocal crypt epithelial cell degeneration (3 males in duodenum and ileum and 1 female in ileum) and minimal to mild dysplasia of the mucosal epithelium (duodenum in 4 males, jejunum in the 1 male with macroscopic watery intestinal content, and ileum in 2 males) with

Downloaded from tpx.sagepub.com at UNIVERSITE DE MONTREAL on September 6, 2015

6

MORTON ET AL.

TOXICOLOGIC PATHOLOGY

FIGURE 1.—(A) Transverse histologic section of normal intestinal villi from the duodenum of a control rat. 20 objective, hematoxylin and eosin (H&E) stain. (B) Transverse histologic section of blunted intestinal villi with dysplastic epithelial cells from the duodenum of a hydroxyureatreated rat. 20 objective, H&E stain. (C) Longitudinal histologic section of normal nonglandular epithelium from the stomach of a control rat. 20 objective, H&E stain. (D) Longitudinal histologic section of dysplastic nonglandular epithelium and submucosal edema from the stomach of a hydroxyurea-treated rat. 20 objective, H&E stain. (E) Longitudinal histologic section of normal bone and bone marrow from the sternum of a control rat. 10 objective, H&E stain. (F) Longitudinal histologic section of decreased cellularity of bone marrow from the sternum of a hydroxyurea-treated rat. 10 objective, H&E stain.

Downloaded from tpx.sagepub.com at UNIVERSITE DE MONTREAL on September 6, 2015

Vol. XX, No. X, 2014

HYDROXYUREA TOXICITY

TABLE 3.—Day 1 mean toxicokinetic values for rats orally administered hydroxyurea. 50 mg/kg AUC (mghr/ml)a

Cmax (mg/ml)

Mean SD

Male

Female

Overall

19.2 2.08

6.91 1.67

13.1 6.74

Male 25.6 5.78

Female

Overall

7.29 1.87

16.5 10.5

500 mg/kg AUC (mghr/ml)a

Cmax (mg/ml)

Mean SD

Male

Female

Overall

183 19.8

180 13.7

182 16.1

Male 887 127

Female

Overall

383 24.6

635 279

1,500 mg/kg AUC (mghr/ml)a

Cmax (mg/ml)

Mean SD

Male

Female

Overall

381 32.3

497 53.1

439 73.8

Male 3,010 774

Female

Overall

2,930 646

2,970 674

Note. Exposures on day 9 at each respective dose level were similar to the exposures on day 1 (data not shown). All means are calculated based on 5 animals per sex per group. Tmax was 1 hr for all animals except for 1 male at 1,500 mg/kg that had a Tmax of 2 hr. AUC ¼ area under the plasma drug concentration curve; Cmax ¼ maximal plasma drug concentration; SD ¼ standard deviation; Tmax ¼ time of maximal plasma drug concentration. a AUC values for individual animals were estimated from time 0 to the last quantifiable time point. AUC intervals were either 0 to 2 or 0 to 4 hr for individuals in the 50-mg/kg dose group; 0 to 24 and 0 to 4 hr for the males and females administered 500 mg/kg/day, respectively; and 0 to 24 hr for the 1,500-mg/kg/day animals.

or without villous atrophy. The dysplasia was characterized by disorganized epithelial cells with variable amounts of cytoplasm and nuclei of various sizes lining blunted villi. In the mesentery of 3 males and 4 females at 500 mg/kg/day, there was minimal to moderate atrophy (lipid depletion) of adipose tissue. The adipocyte atrophy was considered to be secondary to decreased food consumption and the poor clinical condition. Two males had abnormally dark lungs that correlated to multifocal alveolar hemorrhage in 1 animal. Toxicokinetic data in rats are summarized in Table 3. In general, the maximal plasma concentration (Cmax) increased proportionally to dose while area under the plasma drug concentration curve (AUC) increased greater than proportionally to dose. A 30-fold increase in dose (50–1,500 mg/kg/day) resulted in an 180-fold increase in AUC. At 50 mg/kg/day, males had greater exposure than females, but this difference decreased as the dose increased with male and female animals having similar exposures after administration of 1,500 mg/kg on day 1. Exposures at the same dose level were similar on days 1 and 9. No exposure data were available after day 1 for the 1,500mg/kg/day group due to early termination of this group on day 5. In rats, 500 mg/kg/day exceeded the maximum tolerated dose based on failure to gain body weight and microscopic findings in the gastrointestinal tract, thymus, and bone marrow. The dose level of 50 mg/kg was considered the no adverse

7

effect level because the hematologic findings were of limited severity and not adverse, and there were no adverse clinical observations or microscopic findings. At 50 and 500 mg/kg/day, the mean exposure (AUC0–24) in males was 2- to 3-fold the mean exposure in females administered the same dose, and hematologic and microscopic findings were more severe in males. 2-Week Oral Range Finding Study in Dogs Test article–related moribundity occurred in all animals administered 250 or 1,000 mg/kg on day 1 of the study. Within 40 min after dosing, animals administered 1,000 mg/kg had clinical signs of emesis, decreased activity, prostration, inability to rise, blue gums, salivation, and ocular mucus accumulation. Approximately 2 hr later, animals administered 250 mg/kg exhibited clinical signs similar to those noted in animals administered 1,000 mg/kg as well as ataxia and rapid breathing. The heart rates were approximately 200 beats/min and blood oxygen saturation was 70% in animals administered 250 mg/kg; however, methemoglobinemia (see subsequently) may interfere with pulse oximetry results. Due to their moribund condition, animals in both dose groups were euthanized on day 1. Blood samples were collected from those dogs shortly before euthanasia. Values for red blood cells, hemoglobin, and hematocrit were slightly elevated (1.1–1.2  pre-study baseline values in the same animals) in 3 of the 4 dogs administered 250 mg/ kg. Reticulocytes were elevated (2.2–3.0  baseline) and exceeded the historical reference range in the dogs administered 1,000 mg/kg. The elevated erythrocyte mass parameters were consistent with hemoconcentration and shock. Test article–related clinical chemistry findings at 250 mg/kg included elevated glutamate dehydrogenase activity (1.1–2.4  baseline), creatine kinase activity (1.6–2.6  baseline in 2 dogs at 1,000 mg/kg), triglycerides (1.1–2.7  baseline), and glucose (1.2–2.7  baseline) that often exceeded historical reference ranges. Serum phosphorus was decreased (0.56–0.60  baseline) in both animals given 1,000 mg/kg. Changes in triglycerides, glucose, and phosphorus were considered secondary, indirect effects of hydroxyurea related to the moribund condition of the animals. At necropsy, the blood from the animals administered 250 or 1,000 mg/kg was brown and the lungs were diffusely dark. The blue ocular mucous membranes observed prior to death, brown discoloration of the blood at necropsy, and low blood oxygen saturation were consistent with methemoglobinemia; however, methemoglobin was not assessed biochemically. Tissues from the dogs given 250 and 1,000 mg/kg were not examined microscopically. Hydroxyurea was tolerated at 50 mg/kg/day, with no changes in body weight and no test article–related clinical signs. At the end of the study, values for red blood cells, hemoglobin, hematocrit, white blood cells, neutrophils, monocytes, basophils, and platelets were decreased compared with baseline values for the same animals. Values for red blood cells, hemoglobin, and hematocrit were decreased to 0.65 to 0.78  baseline values. Neutrophils were decreased to 0.41 to 0.63  baseline values, while monocytes were decreased

Downloaded from tpx.sagepub.com at UNIVERSITE DE MONTREAL on September 6, 2015

8

MORTON ET AL.

TOXICOLOGIC PATHOLOGY

TABLE 4.—Electrocardiogram data from dogs administered 50-mg/kg/day hydroxyurea orally for 1 month. Incidence Range of increase (beats/minute) Differences between day 23 after dosing compared to day 23 before dosing

Heart rate change (14 beats/min) QTc (10 msec) Differences between day 23 after dosing compared to pre-study baseline values Heart rate change (14 beats/min) QTc (10 msec)

2/6 0/6 3/6 0/6

þ23 to 32 — þ15 to 19 —

Note. QTc ¼ corrected QT interval.

to *0.75  baseline values. At the end of the study, platelets in the male administered 50 mg/kg/day were decreased to 0.66  baseline value and at 165  103/ml were below the lower limit of the historical reference range of 197 to 503  103/ml. No changes were observed for reticulocytes. At the end of the study, creatine kinase activities were increased up to 2.7  baseline values in both dogs given 50 mg/kg/day. There were no other clinical chemistry findings, necropsy findings, or microscopic findings attributed to administration of the hydroxyurea at 50 mg/kg/day. The dose level of 50 mg/kg/day was considered a no adverse effect level because the magnitudes of the clinical pathology changes were modest and not adverse. Due to the early termination of dogs administered 250 mg/kg/day, toxicokinetic analysis was only conducted for dogs administered 50 mg/kg/day. Hydroxyurea was rapidly absorbed with time of maximal hydroxyurea plasma concentration (Tmax) of approximately 1 hr. On day 1, mean Cmax was 40.7 mg/ml and mean AUC was 268 mghr/ml. On day 14, mean Cmax was 39.2 mg/ml and mean AUC was 245 mghr/ml.

observed in most dogs administered hydroxyurea; however, there was no microscopic evidence of cardiac or skeletal muscle injury. Although the cause of elevated creatine kinase activity was not apparent, oxidative damage to muscle caused by free radical metabolites may have contributed. Hydroxyurea did not produce any urinalysis, organ weight, or macroscopic findings. Microscopic findings attributed to the administration of hydroxyurea (summarized in Table 6) included increased overall bone marrow cellularity with decreases in the maturing granulocyte pool and increased brown iron/hemosiderin deposits in bone marrow macrophages and hepatic sinusoidal cells (confirmed with Perl’s iron stain; Figure 2). Toxicokinetic data are summarized in Table 7. Hydroxyurea exposures were similar on days 1 and 27. Peak concentrations of hydroxyurea were achieved within 30 to 60 min of dosing. There were no apparent sex-related differences in exposure or toxicity, and exposures were similar to those observed in the 14-day dog study. Cardiovascular Safety Pharmacology Telemetry Study in Dogs

1-Month Oral Toxicity Study in Dogs Hydroxyurea was tolerated in dogs administered 50 mg/kg/ day orally for 1 month, with no adverse clinical signs or effects on body weights, food consumption, or ophthalmic assessments. ECG findings are summarized in Table 4. When individual predose values were compared to corresponding postdose values on day 23, 2 of the 6 dogs administered hydroxyurea had increased heart rates, but none had significant changes in the QTc. When values recorded after dosing on day 23 were compared to pre-study values for the same dogs, 3 dogs had increased heart rates with no significant changes in QTc. Clinical pathology findings attributed to administration of hydroxyurea at 50 mg/kg/day are summarized in Table 5. Decreases were observed in red cell mass parameters, including red blood cell counts, hemoglobin, and hematocrit. Reticulocytes were increased, consistent with a mild regenerative response. Decreased values compared with baseline values were observed for circulating white blood cells, neutrophils, monocytes, eosinophils, and basophils; however, these decreases were of moderate magnitude. Decreased platelets were seen in both sexes. Hydroxyurea produced slight poikilocytosis in blood smears of 2 males and 2 females, moderate poikilocytosis in 1 male, and target cells in 3 males, while blood cell morphology was normal in all controls. Increased creatine kinase activity that was above the historical reference range was

Cardiac parameters are presented in Figures 3 to 5. Toxicokinetic data for this study are summarized in Table 8. There were no hydroxyurea-related changes obtained for any cardiovascular parameter evaluated at 5 mg/kg. At 15 mg/kg, a test article–related decrease in mean systolic pressure (7 mm Hg) was obtained during the 0.75- to 3.75-HPD period. Oral administration of hydroxyurea at 50 mg/kg resulted in decreases in systolic pressure (8 to 15 mm Hg) and increases in diastolic pressure (þ4 mm Hg) and heart rate (þ22 to þ24 beats/min) during the 0.75- to 9-HPD period. Decreases in QT-interval (16 msec; 0.75–9 HPD) and PR-interval (6 msec; 0.75–3.75 HPD) observed at 50 mg/kg were considered secondary to the increase in heart rate. Decreases in LV maximum systolic pressure (9 mm Hg; 0.75–3.75 HPD) and LV end diastolic pressure (3 mm Hg; 0.75–18 HPD) were obtained along with an increase in LV þdP/dt maximum ranging from 326 to 594 mm Hg/sec during the 0.75- to 18-HPD period. All measures were comparable to vehicle control levels within 24 hr after dose administration. DISCUSSION Hydroxyurea is used in humans and dogs to treat melanoma, myelocytic leukemia, ovarian cancer, head and neck squamous cell carcinoma, polycythemia, and thrombocytosis (FDA

Downloaded from tpx.sagepub.com at UNIVERSITE DE MONTREAL on September 6, 2015

Vol. XX, No. X, 2014

HYDROXYUREA TOXICITY

9

TABLE 5.—Test article–related mean clinical pathology findings in dogs administered hydroxyurea daily by gavage for 1 month at 50 mg/kg/day and the ranges of ratios of individual values near the end of the study relative to baseline values in the same animals.

Red blood cells (106/ml) Ratio compared with baselinea Hemoglobin (g/dl) Ratio compared with baseline Hematocrit (%) Ratio compared with baseline Mean corpuscular volume (fl) Ratio compared with baseline Mean corpuscular hemoglobin (pg) Ratio compared with baseline Mean corpuscular hemoglobin concentration (g/dl) Ratio compared with baseline Red cell distribution width (%) Ratio compared with baseline Reticulocytes (103/ml) Ratio compared with baseline Poikilocytosis incidence Target cells incidence Platelets (103/ml) Ratio compared with baseline White blood cells (103/ml) Ratio compared with baseline Neutrophils (103/ml) Ratio compared with baseline Monocytes (103/ml) Ratio compared with baseline Eosinophils (103/ml) Ratio compared with baseline Basophils (103/ml) Ratio compared with baseline Creatine kinase Ratio compared with baseline

Males N ¼ 3/group

Females N ¼ 3/group

Dose (mg/kg/day)

Dose (mg/kg/day)

0

50

0

50

7.1 0.95–1 17 0.94–1.0 51 0.96–1.01 71 0.97–1.01 23 0.98–1.00 33 0.98–0.99 13 0.96–0.98 41 0.57–1.40 0 0 220 1.15–1.16 6.3 0.76–0.95 3.5 0.73–0.92 0.22 0.42–1.05 0.17 0.95–1.36 0.04 0.40–0.67 181 0.29–1.62

3.6 0.48–0.54 9 0.52–0.57 30 0.57–0.63 84 1.16–1.19 25 1.06–1.07 30 0.90–0.92 19 1.28–1.51 111 1.32–4.0 3 3 165 0.50–0.75 4.1 0.40–0.48 2.0 0.27–0.42 0.15 0.18–0.40 0.05 0.24–0.33 0.01 0.06–0.20 464 0.78–1.85

7.5 0.99–1.14 18 0.99–1.17 54 1.01–1.14 72 0.99–1.02 24 0.99–1.02 33 0.99–1.02 13 0.96–0.98 50 1.06–3.3 0 0 319 0.85–0.98 6.6 0.66–0.94 3.8 0.59–0.89 0.21 0.55–0.93 0.26 1.35–3.9 0.06 0.56–1.43 185 0.93–1.04

3.6 0.46–0.58 9 0.50–0.61 30 0.55–0.67 84 1.14–1.20 25 1.06–1.09 30 0.90–0.92 19 1.36–1.56 95 1.23–4.7 2 0 200 0.52–0.64 3.6 0.37–0.50 1.6 0.10–0.44 0.12 0.21–0.74 0.05 0.18–0.67 0.01 0–0.50 362 0.63–3.5

Note. No hydroxyurea-related findings were observed for mean platelet volume, lymphocytes, activated partial thromboplastin time, prothrombin time, fibrinogen, alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase, g-glutamyltransferase, total protein, albumin, globulin, albumin/globulin ratio, total bilirubin, glucose, cholesterol, blood urea nitrogen, creatinine, calcium, phosphorus, sodium, potassium, chloride, and urinary parameters including appearance, pH, protein, ketones, bilirubin, urobilinogen, glucose, specific gravity, and formed elements in sediment. a Range of values (ratios) when individual values at the end of the study are divided by the baseline values for the same animal and the same parameter.

TABLE 6.—Microscopic findings in dogs caused by hydroxyurea given daily by gavage for 1 month at 50 mg/kg/day. Males N¼3 Bone marrow Increased cellularity Present 3 Decreased mature granulocyte precursors Mild 1 Moderate 1 Marked 1 Increased iron pigment in macrophages Minimal — Mild 3 Liver Increased iron pigment in sinusoidal cells Minimal 2 a

Denotes no test article–related finding.

Females N¼3

3 —a — 3 2 1

3

2010). In humans with sickle-cell disease, hydroxyurea is the only treatment option to reduce the frequency and severity of vaso-occlusive crises (Bunn 1997). Hydroxyurea has also been used to treat mast cell neoplasms and hypereosinophilic syndrome in dogs (Marconato, Bonfanti, and Fileccia 2007; Rassnick et al. 2010). Although hydroxyurea is generally tolerated in patients, the margin of safety is low due to myelosuppression. Repeated assessment of circulating erythrocytes, leukocytes, and platelets is recommended during therapy (FDA 2012). The most common and earliest adverse finding in human patients is leukopenia, followed by anemia and thrombocytopenia (FDA 2010, 2012; Rassnick et al. 2010). Most available toxicity data from animals are found only in product labels, are incomplete, and are often described only in summary statements (reviewed by Liebelt et al. 2007). Available information is often insufficient to design optimal nonclinical

Downloaded from tpx.sagepub.com at UNIVERSITE DE MONTREAL on September 6, 2015

10

MORTON ET AL.

TOXICOLOGIC PATHOLOGY

FIGURE 2.—(A) Bone marrow from a control female dog. 40 objective, hematoxylin and eosin (H&E) stain. (B) Bone marrow from a female dog administered hydroxyurea at 50 mg/kg/day for 30 days. There is increased cellularity with almost no fat in the marrow cavity, very few maturing granulocytes, and increased large, immature, hematopoietic cells compared to the control. Brown iron deposits are much more common than in control bone marrow. 40 objective, H&E stain. (C) Brown iron pigment in many sinusoidal cells of the liver. 40 objective, H&E stain. (D) Brown pigment in hepatic sinusoidal cells stains positively with Perl’s iron stain. 40 objective. TABLE 7.—Mean toxicokinetic data in dogs administered hydroxyurea at 50 mg/kg/day for 1 month.

Cmax (mg/ml) AUC0–24 hr (mghr/ml) Tmax (h)

Males

Females

N¼3

N¼3

Day 1

Day 27

Day 1

Day 27

55.8 241 0.5

58.4 218 0.7

54.6 189 0.5

62.0 191 0.5

Note. AUC ¼ area under the plasma drug concentration curve; Cmax ¼ maximal plasma drug concentration; Tmax ¼ time of maximal plasma drug concentration.

studies to assess toxicity of other pharmaceutical candidates in combination with hydroxyurea or to critically compare the toxicity profiles of hydroxyurea and alternative therapies. The studies described in this article supplement the existing literature and provide additional details regarding the

toxicity of hydroxyurea in rats and dogs. The hydroxyurearelated decreased granulocytes, erythrocytes, and platelets reported in humans, canine patients, and previous nonclinical studies were confirmed in the studies presented here (Lerner et al. 1966; Rassnick et al. 2010). Red blood cells, hemoglobin, hematocrit, neutrophils, and monocytes showed similar decreases in the 14-day and 1-month dog studies, with lower values generally observed at 1 month than at 14 days. Platelets were decreased in males at 14 days and in both sexes after 1 month. Mean corpuscular volume was increased in dogs after 1 month of hydroxyurea administration, but these changes were not observed in the 14-day study. Increased mean corpuscular volume was accompanied by slightly increased mean corpuscular hemoglobin and slightly decreased mean corpuscular hemoglobin concentration. Paradoxically, in the 1-month dog study (but not in the 2 dogs administered hydroxyurea at 50 mg/kg/day for 14 days), the bone marrow demonstrated increased cellularity and reticulocytes were increased,

Downloaded from tpx.sagepub.com at UNIVERSITE DE MONTREAL on September 6, 2015

Vol. XX, No. X, 2014

HYDROXYUREA TOXICITY

11

FIGURE 3.—Time course graphs illustrating mean heart rate (top left), systolic pressure (top right), diastolic pressure (bottom left), and heart rate (bottom right) measures following treatment at 0, 5, 15, and 50 mg/kg hydroxyurea to male Beagle dogs (N ¼ 4). Symbols represent mean values for 15 preceding minutes.

indicating a regenerative response to bone marrow toxicity. Internal data (data not shown) estimate the half-life (t1/2) of hydroxyurea to be *1.5 hr in rats and *2.5 hr in dogs following a single intravenous (IV) dose, while others estimate t1/2 in the rat at 35 to 65 min (Philips et al. 1967). The human t1/2 is about 3.5 hr (Liebelt et al. 2007). Since animals were dosed just once daily and hydroxyurea has a short t1/2, the decreases in circulating blood cells with increased bone marrow cellularity may have resulted from transient daily bone marrow toxicity at peak concentrations of hydroxyurea followed by a daily period of recovery and regeneration as hydroxyurea in tissues decreased below pharmacologically active concentrations. Philips et al. (1967) reported that cellular damage in intestinal and bone marrow cells of rats was maximal at 2 to 3 hr after administration of a single high dose of hydroxyurea (1,840 mg/kg IV), was diminishing by 5 hr, and largely resolved within 40 hr. Similarly, cellular damage in lymphoid organs had resolved within 40 hr after administration of hydroxyurea. Although Philips et al. (1967) reported that there was no depletion of cellularity in lymph nodes, spleen, and thymus following a single dose, rats treated for 2 weeks at 500 mg/kg/day had decreased cellularity of the

thymus, lymph nodes, and bone marrow, indicating that damage can be cumulative with repeated daily administration. The toxicities observed in rats and dogs can be explained in most cases by the published mechanisms of action of hydroxyurea. Hydroxyurea is rapidly metabolized to a carbomoyl nitroso intermediate and then to nitroxide compounds (including nitric oxide) that are responsible for physiologic and toxic effects (Koviac 2011; Yarbro 1992). These nitroxides act as free radicals, interfering with electron transfer, facilitating formation of reactive oxygen species inducing oxidative stress, interacting with nucleic acids and proteins to impair cellular functions, and altering cell signaling (Koviac 2011). Hydroxyurea is mutagenic and clastogenic in multiple assays, although not all reports agree (reviewed by Liebelt et al. 2007). The free radical nitroxide inactivates ribonucleotide reductase by quenching a tyrosyl free radical and chelating the iron in the active site in the M2 subunit, selectively preventing incorporation of thymidine into deoxyribonucleic acid (DNA), halting DNA synthesis, and causing death of dividing cells (Philips et al. 1967). Only cells in S phase that are incorporating thymidine into new DNA are killed. Synthesis of ribonucleic acid and protein are not affected (FDA 2010). DNA

Downloaded from tpx.sagepub.com at UNIVERSITE DE MONTREAL on September 6, 2015

12

MORTON ET AL.

TOXICOLOGIC PATHOLOGY

FIGURE 4.—Time course graphs illustrating mean PR interval (top left), QRS interval (top right), QT interval (bottom left), and QTc interval (bottom right) measures following treatment at 0, 5, 15, and 50 mg/kg hydroxyurea to male Beagle dogs (N ¼ 4). Symbols represent mean values for 15 preceding minutes.

repair is also inhibited by hydroxyurea (Koviac 2011; Yarbro 1992). This selective inhibition of DNA synthesis and cell division in rapidly dividing cells in S phase may partially explain the antineoplastic activity and toxic effects of hydroxyurea in rapidly dividing tissues, including bone marrow, gastrointestinal tract, thymus, and skin. Modeling of clinical pharmacokinetic data from hydroxyureatreated cancer patients (Tracewell et al. 1995) for projected human exposures at 35 mg/kg, the maximum recommended clinical dose for treatment of sickle-cell disease, resulted in projected Cmax and AUC values of 31 mg/ml and 58 mghr/ml, respectively. Exposure multiples comparing the toxicokinetic values obtained in these toxicology studies in rats and dogs to the projected human exposure demonstrate low multiples for hydroxyurea toxicity in dogs and rats (Table 9). In the 1-month dog study, exposure margins at the no observed adverse effect level (NOAEL) of 50 mg/kg/day were 1.9 and 3.5, respectively, the projected human clinical Cmax of 31 mg/ml and AUC of 58 mghr/ml. In the rat study, the exposure margins at the NOAEL of 50-mg/kg dose were 0.4 and 0.3 the projected clinical Cmax and AUC, respectively. At 500 mg/kg in rats, a dose that was not tolerated based on the failure to gain body

weight and microscopic findings in the gastrointestinal tract, thymus, and bone marrow, the Cmax and AUC margins were 6.9 and 11, respectively. Hydroxyurea metabolites including nitric oxide react with hemoglobin to form methemoglobin and nitric oxide-hemoglobin (Jiang et al. 1997; Rupon et al. 2000). These altered forms of hemoglobin cannot bind and release oxygen. Acute, clinically relevant methemoglobinemia was described in a dog that accidentally ingested an estimated 200 mg/kg of hydroxyurea (Bates 2008; Wray 2008). In that dog, prostration, cyanosis, tachycardia, dark brown blood, and increased creatine kinase activity were observed. These findings were similar to the acute toxicity in dogs reported here. Hypoxia secondary to methemoglobinemia was the likely cause of acute clinical signs and moribundity in dogs administered a single dose of hydroxyurea at 250 mg/kg. In addition to decreased bone marrow production of erythrocytes and reactions with hemoglobin to form methemoglobin, hydroxyurea produces other effects on erythrocytes. Hydroxyurea promotes the production of fetal hemoglobin through activation of soluble guanylate cyclase, and this increased fetal hemoglobin is beneficial in patients with

Downloaded from tpx.sagepub.com at UNIVERSITE DE MONTREAL on September 6, 2015

Vol. XX, No. X, 2014

HYDROXYUREA TOXICITY

13

FIGURE 5.—Time course graphs illustrating mean left ventricular (LV) maximum systolic pressure (top left), LV end diastolic pressure (top right), LV change in blood pressure over time (þdP/dt) max (bottom left), and activity (bottom right) following treatment with vehicle, 5-, 15-, and 50mg/kg hydroxyurea to male Beagle dogs (N ¼ 4). Symbols represent mean value for 15 preceding minutes. TABLE 8.—Toxicokinetic summary for dogs administered hydroxyurea in the cardiovascular safety pharmacology study. Dose (mg/kg) 5 15 50 50 (Toxicokinetic phase)

4 HPD (mg/ml)

Cmax (mg/ml)

Tmax (HPD)

1.73 5.61 23.6 24.5

— — — 65.8

— — — 0.6

Note. Cmax ¼ maximal plasma drug concentration; HPD ¼ hours postdose; Tmax ¼ time of maximal plasma drug concentration.

sickle-cell disease because fetal hemoglobin reduces the formation of dense aggregates of hemoglobin S that deform erythrocytes and contribute to vaso-occlusive disease (Bunn 1997; King 2004; Orringer et al. 1991). Hydroxyurea also increases the erythrocyte membrane surface area and mean

cell volume (Charache et al. 1995; Engstro¨m and Lo¨fvenberg 1998; Orringer et al. 1991). Orringer et al. (1991) state that hydroxyurea causes higher mean corpuscular hemoglobin in dogs there by drawing in more intracellular water, but that similar increases in humans are not caused by the same mechanism. Increased mean corpuscular volume in patients with sicklecell disease has been attributed to increased intracellular hemoglobin, perhaps with additional contributions by alterations of erythrocyte membranes and other factors (Charache et al. 1995). Nitric oxide formed as a metabolite of hydroxyurea probably was the major mediator of cardiovascular effects (including decreased systolic BP, increased diastolic BP, and increased heart rate) observed in dogs in these studies (Jiang et al. 1997; King 2004). Nitric oxide is a potent vasodilator. Increased nitric oxide formation results in smooth muscle

Downloaded from tpx.sagepub.com at UNIVERSITE DE MONTREAL on September 6, 2015

14

MORTON ET AL.

TOXICOLOGIC PATHOLOGY

TABLE 9.—Estimated human exposure margins for rat and dog toxicokinetic data based on projected clinical exposures at 35 mg/kg/day.

Dog NOAELb Rat NOAELc Rat LOAELc

Dose (mg/kg)

Cmax (mg/ml)

AUC (mghr/ml)

Cmax exposure margina

AUC exposure margina

50 50 500

60.2 13.1 182

204.5 16.5 635

1.9 0.4 6.9

3.5 0.3 11

Note. AUC ¼ area under the plasma drug concentration curve; Cmax ¼ maximal plasma drug concentration; NOAEL ¼ no observed adverse effect level; LOAEL ¼ lowest observed adverse effect level. a Projected Cmax value of 31 mg/ml and AUC value of 58 mghr/ml were used in this calculation. b Data used from day 27 toxicokinetic samples from the 1-month dog study. c Rat exposures were averaged for male and female animals.

relaxation, decreased BP, and compensatory increased heart rate (Lakshmana, Hargis, and Woodcock 2012). Nitric oxide stimulates cyclic guanosine monophosphate (cGMP) production that in turn regulates multiple cardiac ion channels. Increased levels of cGMP act to promote parasympathetic activity and inhibit b-adrenergic stimulation of the heart, resulting in negative inotropic effects, decreased cardiac contractility, decreased systolic BP, and compensatory increased heart rate (Fischmeister et al. 2005). There were no hydroxyurea-related microscopic heart findings resulting from the functional changes. While the acute hypoxia, blue mucous membranes, prostration and shock leading to euthanasia, and brown blood observed at necropsy in dogs administered 250 mg/kg were attributed to methemoglobinemia, functional cardiac changes resulting from increased nitric oxide in tissues may have further compromised the ability of the dogs to compensate for effects of methemoglobinemia. In conclusion, hydroxyurea produced evidence of methemoglobinemia, prostration, and death in dogs after a single dose of 250 mg/kg, and was poorly tolerated in rats at 500 mg/kg/day. The only findings in rats related to hydroxyurea administration at 50 mg/kg/day were lower circulating leukocytes, whereas at 500 mg/kg/day findings also included lower red cell mass and platelets, decreased cellularity of bone marrow, thymus, and lymph nodes, epithelial degeneration in the crypts of the small intestine, and gastric and intestinal epithelial dysplasia. In dogs administered 50 mg/kg/day for 1 month, findings included decreased circulating erythrocytes, leukocytes, and platelets, increased bone marrow cellularity, and elevated creatine kinase activity. Cardiovascular findings in dogs consisted of decreased systolic BP, increased diastolic BP, and increased heart rate, and were consistent with increased formation of nitric oxide from hydroxyurea. AUTHOR CONTRIBUTION Daniel Morton, Lori Reed, Wenhu Huang, John M. Marcek, Robert Austin-LaFrance, Carrie A. Northcott, Scott H. Schelling, Bradley E. Enerson, and Lindsay Tomlinson contributed to conception or design, data acquisition, analysis, or interpretation; Daniel Morton, Robert Austin-LaFrance, Carrie A. Northcott, and John M. Marcek drafted the manuscript; Daniel Morton, Lori Reed, Wenhu Huang, John M. Marcek, Robert Austin-LaFrance, Carrie A. Northcott, Scott H. Schelling, Bradley E. Enerson, and Lindsay Tomlinson critically revised

the manuscript; and Bradley E. Enerson gave final approval. All authors agreed to be accountable for all aspects of work in ensuring that questions relating to the accuracy or integrity of any part of the work are appropriately investigated and resolved. REFERENCES Bates, N. (2008). Hydroxycarbamide (hydroxyurea) toxicity in dogs. J Small Animal Pract 49, 216. Bunn, H. F. (1997). Pathogenesis and treatment of sickle cell disease. N Eng J Med 337, 762–69. Charache, S., Terrin, M. L., Moore, R. D., Dover, G. J., Barton, F. B., Eckert, S. V., McMahon, R. P., and Bonds, D. R. (1995). Effect of hydroxyurea on the frequency of painful crises in sickle cell anemia. N Eng J Med 332, 1317–22. Engstro¨m, K., and Lo¨fvenberg, E. (1998). Treatment of myeloproliferative disorders with hydroxyurea: Effects on red blood cell geometry and deformability. Blood 91, 3986–91. Fischmeister, R., Castro, L., Abi-Gerges, A., Rochais, F., and Vandecasteele, G. (2005). Species- and tissue-dependent effects of NO and cyclic GMP on cardiac ion channels. Comp Biochem Physiol Part A: Mol Integrative Physiol 142, 136–43. Food and Drug Administration, Center for Drug Evaluation and Research. (2010). Hydrea1 packet insert. Accessed February 17, 2014. http://www .accessdata.fda.gov/drugsatfda_docs/label/2010/016295s040lbl.pdf. Food and Drug Administration, Center for Drug Evaluation and Research. (2012). Droxia1 packet insert. Accessed February 17, 2014. http://www .accessdata.fda.gov/drugsatfda_docs/label/2012/016295s041s042lbl.pdf. Friedrich, S., Raff, K., Landthaler, M., and Karrer, S. (2004). Cutaneous ulcerations on hands and heels secondary to long-term hydroxyurea treatment. Eur J Dermatol 14, 343–6. Italia, K., Colah, R., and Ghosh, K. (2013). Hydroxyurea could be a good clinically relevant iron chelator. PLOS One 8, 1–5. Jiang, J., Jordon, S. J., Barr, D. P., Gunther, M. R., Maeda, H., and Mason, R. P. (1997). In vivo production of nitric oxide in rats after administration of hydroxyurea. Molec Pharmacol 52, 1081–86. Juul, T., Malolepszy, A., Dybkaer, K., Kidmose, R., Rasmussen, J. T., Andersen, G. R., Johnsen, H. E., Jorgensen, J.-E., and Andersen, S. U. (2010). The in vivo toxicity of hydroxyurea depends on its direct target catalase. J Biol Chem 285, 21411–415. King, S. B. (2004). Nitric oxide production from hydroxyurea. Free Radical Biol Med 37, 737–44. Koviac, P. (2011). Hydroxyurea (therapeutics and mechanism): Metabolism, carbamoyl nitroso, nitroxyl, radicals, cell signaling and clinical applications. Med Hypotheses 76, 24–31. Lakshmana, S., Hargis, J. C., and Woodcock, H. L. (2012). Unlocking the binding and reaction mechanism of hydroxyurea substrates as biological nitric oxide donors. J Chem Inf Modeling 52, 1288–97. Lerner, L. J., Bianchi, A., Yiacas, E., and Borman, A. (1966). Effects of hydroxyurea and related compounds on the blood and marrow of experimental animals. Canc Res 25, 2292–96.

Downloaded from tpx.sagepub.com at UNIVERSITE DE MONTREAL on September 6, 2015

Vol. XX, No. X, 2014

HYDROXYUREA TOXICITY

Liebelt, E. L., Balk, S. J., Faber, W., Fisher, J. W., Hughes, C. L., Lanzkron, S. M., Lewis, K. M., Marchetti, F., Mehendale, H. M., Rogers, J. M., Shad, A. T., Skalko, R. G., and Stanek, E. J. (2007). NTP-CERHR expert panel report on the reproductive and developmental toxicity of hydroxyurea. Birth Defects Res (Part B) 80, 259–366. Marconato, L., Bonfanti, U., and Fileccia, I. (2007). Unusual dermatologic toxicity of hydroxyurea in two dogs with spontaneously occurring tumours. J Sm An Pract 48, 514–17. O’Branski, E. E., Ware, R. E., Prose, N. S., and Kinney, T. R. (2001). Skin and nail changes in children with sickle cell anemia receiving hydroxyurea therapy. J Am Acad Dermatol 44, 859–61. Orringer, E. P., Blythe, D. S. B., Johnson, A. E., Phillips, G., Dover, G. J., and Parker, J. C. (1991). Effects of hydroxyurea on hemoglobin F and water content in the red blood cells of dogs and of patients with sickle cell anemia. Blood 78, 212–16. Philips, F. S., Sternberg, S. S., Schwartz, H. S., Cronin, A. P., Sodergren, J. E., and Vidal, P. M. (1967). Hydroxyurea I. Acute cell death in proliferating tissues in rats. Cancer Res 27, 61–74. Rassnick, K. M., Al-Sarraf, R., Bailey, D. B., Chretin, J. D., Phillips, B., and Zwhalen, C. H. (2010). Phase II open-label study of single-agent hydroxyurea for treatment of mast cell tumours in dogs. Vet Comp Oncol 8, 103–11.

15

Rupon, J. W., Domingo, S. R., Smith, S. V., Gummadi, B. K., Shields, H., Ballas, S. K., King, S. B., and Kim-Shapiro, D. B. (2000). The reactions of myoglobin, normal adult hemoglobin, sickle cell hemoglobin and hemin with hydroxyurea. Biophys Chem 84, 1–11. Silva-Pinto, A. C., Dias-Carlos, C., Saldanha-Araujo, F., Ferreira, F. I. S., Palma, P. V. B., Araujo, A. G., Queiroz, R. H. C., Elion, J., Covas, D. T., Zago, M. A., and Panepucci, R. A. (2014). Hydroxycarbamide modulates components involved in the regulation of adenosine levels in blood cells from sickle-cell anemia patients. Ann Hematol 93, 1457–65. Tracewell, W. G., Trump, D. L., Vaughan, W. P., Smith, D. C., and Gwilt, P. R. (1995). Population pharmacokinetics of hydroxyurea in cancer patients. Cancer Chemother Pharmacol 35, 417–22. Venkayala, S. L., Hargis, J. C., and Woodcock, H. L. (2012). Unlocking the binding and reaction mechanism of hydroxyurea substrates as biological nitric oxide donors. J Chem Info Modeling 52, 1288–97. Vassallo, C., Passamonti, F., Adrigo, M., Nolli, G., Mangiacavalli, S., and Boroni, G. (2001). Muco-cutaneous changes during long-term therapy with hydroxyurea in chronic myeloid leukemia. Clin Exp Dermatol 26, 141–48. Wray, J. D. (2008). Methaemoglobinaemia caused by hydroxycarbamide (hydroxyurea) ingestion in a dog. J Small Animal Pract 49, 211–15. Yarbro, J. W. (1992). Mechanisms of action of hydroxyurea. Semin Oncol 19, 1–10.

For reprints and permissions queries, please visit SAGE’s Web site at http://www.sagepub.com/journalsPermissions.nav.

Downloaded from tpx.sagepub.com at UNIVERSITE DE MONTREAL on September 6, 2015