Adverse side effects, including agranulocytosis and anemia, from methimazole treatment of a hyperthyroid cat

Abstract

A 13-year-old spayed female domestic shorthair cat was brought to a veterinary clinic because of lethargy and anorexia. The cat had a history of hyperthyroidism and had been treated with methimazole for 40 d before the appointment. Hematologic examination revealed a severe non-regenerative anemia and leukopenia characterized by severe neutropenia along with low basophil and eosinophil numbers. Discontinuation of methimazole resulted in the complete return to normal values of the white blood cell numbers and a marked increase in red blood cell numbers. This case demonstrated the need for follow-up veterinary examinations after prescription of methimazole treatment.

CASE DESCRIPTION

A 13-year-old spayed female domestic shorthair indoor cat was brought to a veterinary clinic as an emergency case because of lethargy, anorexia, and weakness. In this report, the day (D) of presentation will be referred to as D1 and all other days will be in reference to it. On D40, the cat had been diagnosed with hyperthyroidism (HT) based on clinical findings — body weight (BW) of 2.5 kg, body condition score (BCS) of 3/9 [reference interval (RI): 5/9], cachexia, polyphagia, hyperactivity — and a total thyroxine concentration (TT4) of 257 nmol/L (RI: 10 to 60 nmol/L). The HT was managed with methimazole (Tapazole 5 mg tablets; Paladin, Montreal, Quebec), 1 mg/kg BW, q12h, PO. The 1st recheck examination for the cat was on D27, at which time the clinical signs had resolved and the cat was euthyroid (TT4: 52 nmol/L). Other bloodwork to check overall health and acclimation to methimazole was declined at that time.

At the emergency visit, the cat was lethargic, weak, and less fractious than before. The abnormal findings included a skin tent and capillary refill time > 2 s, rectal temperature of 36.5°C (RI: 38 to 39.4°C), cachexia, and a new, Grade-2/5 heart murmur. The BW was 3.2 kg. The heart murmur was thought to be hemic, secondary to the possible anemia. The continuing low BCS was likely long-standing from the previous HT. A complete blood (cell) count (CBC), PCV, biochemistry panel, and urinalysis were obtained.

The CBC revealed RBCs 3.53 × 10¹²/L (RI: 6.54 to 12.2 × 10¹²/L), hematocrit 12.3% (RI: 30.3 to 52.3%), HGB 4.6 g/dL (RI: 9.8 to 16.2 g/dL), mean corpuscular volume 34.8 fL (35.9 to 53.1 fL), mean corpuscular hemoglobin concentration 37.4 g/dL (RI: 28.1 to 35.8 g/dL), WBCs 2.86 × 10⁹/L (RI: 2.87 to 17.02 × 10⁹/L), neutrophils 0.16 × 10⁹/L (RI: 2.3 to 10.29 × 10⁹/L), monocytes 1.46 × 10⁹/L (RI: 0.05 to 0.67 × 10⁹/L), eosinophils 0.01 × 10⁹/L (RI: 0.17 to 1.57 × 10⁹/L), and basophils 0.00 × 10⁹/L (RI: 0.01 to 0.26 × 10⁹/L). The PCV was 14%. These findings indicated a severe non-/pre-regenerative microcytic hyperchromic anemia and mild leukopenia characterized by a severe neutropenia accompanied by a mild monocytosis. Blood-smear evaluation did not show any abnormalities, including no microagglutination. The biochemistry values showed glucose 17.4 mmol/L (RI: 3.95 to 8.84 mmol/L), symmetrical dimethylarginine (SDMA) 21 μg/dL (RI: 0 to 14 μg/dL), and urea 35.6 mmol/L (RI: 5.7 to 12.9 mmol/L). The urine sample was orange and clear, with specific gravity 1.050 (RI: > 1.035), and the dipstick revealed 1+ protein (RI: 0), 2+ glucose (RI: 0), pH 6 (RI: 5.5 to 7.5), and 2+ blood (RI: 0).

Treatment with methimazole was discontinued to see if the clinical signs improved. On D2, the cat was taken to a referral center (Western College of Veterinary Medicine, Saskatoon, Saskatchewan). On arrival, the PCV was 23%, a 9% increase in the 24 h since discontinuation of methimazole. On D3, the PCV was retested at 24% and clinical signs had resolved. At the referral center, the cat was tested for FeLV/FIV with a negative result, most likely ruling out those agents as causes for secondary immune-mediated hemolytic anemia (IMHA). Due to the severe anemia, not enough blood was taken to test for Mycoplasma haemofelis. Instead, empirical treatment was started with a 100 mg/mL suspension of doxycycline (Western College of Veterinary Medicine), 0.3 mL (10 mg/kg), q12h for 14 d. Based on the marked improvement in the cat’s anemia and the negative infectious disease tests, it was concluded that the likely cause of the anemia was methimazole toxicity.

After receiving doxycycline for 5 d, the cat’s anorexia returned, prompting the owners to stop antibiotic treatment. Throughout the week the cat developed new clinical signs, hypodipsia and occasional episodes of vomiting, and became lethargic again. On D9, the owners noted that the cat had become icteric. Follow-up bloodwork (CBC, biochemistry, electrolytes, SDMA, and TT4) and urinalysis were obtained. On physical examination, it was noted that the cat’s BW had dropped 0.5 kg, BCS remained at 3/9, the sclera and mucous membranes were icteric, there was a prolonged skin tent, and menace response was absent. The heart murmur had resolved. The bloodwork showed RBCs 5.66 × 10¹²/L (RI: 6.54 to 12.2 × 10¹²/L), hematocrit 22.3% (RI: 30.3 to 52.3%), HGB 7.6 g/dL (RI: 9.8 to 16.2 g/dL), reticulocytes 96.8 K/μL, WBCs 20.94 × 10⁹/L (RI: 2.87 to 17.02 × 10⁹/L), neutrophils 17.06 × 10⁹/L (RI: 2.3 to 10.29 × 10⁹/L), monocytes 0.96 × 10⁹/L (RI: 0.05 to 0.67 × 10⁹/L), and platelets 606 K/μL (RI: 151 to 600 K/μL). No microagglutination was seen and no poikilocytes were evident on blood smear. The CBC showed that the agranulocytosis (low neutrophils, basophils, and eosinophils) had completely resolved with the neutrophil count now elevated. Furthermore, the anemia, which had been non-regenerative at the initial emergency visit, was regenerative. These findings suggested that the cat was recovering from myelosuppression. The initial insult to the bone marrow was likely caused by methimazole, as the problem improved solely with discontinuation of the drug.

The cat’s biochemistry, electrolytes, and SDMA panel results showed ALT 715 U/L (RI: 12 to 130 U/L), ALP 485 U/L (RI: 14 to 111 U/L), GGT 14 U/L (RI: 0 to 4 U/L), total bilirubin 287 μmol/L (RI: 0 to 15 μmol/L), cholesterol 6.49 mmol/L (RI: 1.68 to 5.81 mmol/L), TT4 > 257 mmol/L (RI: 10 to 60 mmol/L), and SDMA 18 μg/dL (RI: 0 to 14 μg/dL). The urine sample was orange and clear, urine specific gravity was 1.050 (RI: > 1.035), and the urine dipstick revealed 1+ protein (RI: 0), pH 6 (RI: 5.5 to 7.5), and 2+ blood (RI: 0). The cat’s hyperglycemia and glucosuria had resolved after the initial emergency visit but it had developed a hypercholesterolemia. These changes were most likely due to the return of the anorexia and relapsed HT consequently increasing metabolism and depleting glycogen stores. The urea also returned to normal, most likely due to the prolonged period of anorexia.

New findings were elevations in liver enzymes and hyperbilirubinemia. These changes, with the exception of an elevated GGT, were consistent with methimazole toxicity (1). Interestingly, the icterus and biochemical changes were noted on D9 and D10, respectively, and thus occurred after methimazole was discontinued. Since neither the timeline nor the GGT elevation was consistent with methimazole toxicity, it was important to consider more likely differential diagnoses. For example, it was possible, that due to the period of severe neutropenia, and immunosuppression, the cat developed cholangiohepatitis. Because of the period of prolonged anorexia, it was possible that the cat developed hepatic lipidosis, further compromising the liver. Without cholecystocentesis or liver aspirates/biopsy, it was impossible to rule out these other causes. After discontinuation of methimazole, the cat became HT again, which could also have caused mild increases in ALP or ALT. The cat’s clinical signs of anorexia, lethargy, and vomiting were not typical of HT and were likely secondary to the hyperbilirubinemia and compounded by the anemia. The hyperbilirubinemia could have also explained the neurologic signs. Otherwise, the lack of menace response could have been due to retinal detachment from hypertension due to HT, but this could not be confirmed as a blood pressure was not measured.

The practice was not equipped to further investigate the cat’s liver changes. The decision was made to wait to see if signs would resolve with discontinuation of methimazole. To manage vomiting, the cat was given an injection of 10 mg/mL maropitant (Cerenia; Zoetis, Kirkland, Quebec), 1 mg/kg, SC. For treating the cat at home, the owners were given oral maropitant (Cerenia, 4 mg tablets; Zoetis), q24h for 4 d. To resolve the dehydration, the cat was given NaCl 0.9%, 120 mL, SC. Since the cat did not improve by D13, the owners considered euthanasia. The cat was euthanized on D20. At the time of euthanasia, icterus had decreased and PCV had increased to 33%. The continuous improvement in RBC without treatment further supported the notion that the anemia was due to methimazole administration.

DISCUSSION

In addition to demonstrating a well-known side effect of oral methimazole administration, agranulocytosis, this case report helps evidence a less-documented potential side effect: anemia. In 1988, Peterson et al evaluated 262 HT cats treated with 10 to 15 mg of oral methimazole for 3 y (1). Within the first 1 to 2 mo after starting treatment, 4.7% of cats developed a temporary leukopenia (1). In addition, within the first 3 mo, 1.5% of cats developed a severe agranulocytosis that did not resolve until 1 wk after methimazole was discontinued (1). Sartor et al showed that 0.12% of cats receiving oral methimazole and 0.07% of cats receiving transdermal methimazole, at a dosage of 2.5 mg q12h, developed neutropenia (2). Neutropenia secondary to methimazole was also identified in 1 of 16 cats with aplastic anemia in a retrospective study evaluating 128 feline bone marrow sample reports (3). Similar to cats in those previous studies, the cat discussed in this case report received 2.5 mg of oral methimazole q12h and developed a severe neutropenia along with low eosinophil and basophil counts. After methimazole treatment was stopped and no other treatment for immune compromise was pursued, the cat’s neutrophil count increased to above normal range. The eosinophil and basophil values also returned to within reference intervals.

Although methimazole is cited as a cause of secondary IMHA or aplastic anemia in cats in literature reviews on immune-mediated hematologic diseases, this adverse effect is not well-documented in the literature (4,5). Interestingly, anemia was not noted in the largest- and longest-scale study investigating methimazole toxicity, by Peterson et al (1). The authors did determine that 1.9% of cats in a subgroup of 160 that underwent direct antiglobulin testing (Coomb’s test) to screen for IMHA tested positive; however, they did not develop anemia (1). Similarly, lupus, another autoimmune condition linked to hemolytic anemia, was screened by testing serum antinuclear antibodies in a subgroup of 238 cats (1). Despite 21.8% of cats testing positive, none of the cats went on to develop anemia (1). This raises a question: in the absence of anemia, do these tests have any value in suggesting the role of methimazole in IMHA? A 2018 study investigating the validity of a Coomb’s test and antinuclear antibody titres in 89 healthy cats determined it was possible to obtain positive test results in healthy animals using various tests available commercially (6). Thus, since the animals in the Peterson et al study were healthy, it is not possible to draw any conclusions from these positive test results (1,6).

The documented association between methimazole and anemia was in 1 cat, in a retrospective study by Weiss (3). Although no cats in the study by Sartor et al developed anemia, cats in the oral methimazole group had a gradual decline in hematocrit (40.5 to 37%), whereas cats in the transdermal group remained above 41% (2). This case report demonstrates the development of a severe anemia (PCV 14%) due to methimazole 40 d after initiating treatment. This was suspected since discontinuing methimazole was associated with an increase in PCV of 9% in 1 d, 10% in 2 d, and 19% in 20 d.

In the case reported here, the pathogenesis through which methimazole caused anemia was thought to be the same as in the report by Weiss: bone marrow suppression (3). Aspects of the report that supported this were the finding of multiple cytopenias, the change in anemia from non-regenerative to regenerative after discontinuation of methimazole (suggesting bone marrow recovery), and the lack of evidence to support autoimmune destruction of RBCs (3,7). Without a bone marrow biopsy, it is impossible to know if this theory is true. Contrary to this theory, 1 case report showed that carbimazole, a metabolite of methimazole, causes anemia via secondary IMHA (8).

One of the main limitations of this case report was the inability to rule out all other causes of anemia. Secondary IMHA in cats is more common than primary and can be a result of FeLV/FIV infection, Mycoplasma infection, neoplasia, or toxicity from a drug such as methimazole. Although this cat had a negative FeLV/FIV SNAP test, empirical treatment for M. haemofelis was discontinued prematurely due to anorexia. Furthermore, because no imaging was done, neoplasia and chronic inflammatory conditions could not be ruled out. Organ dysfunctions such as chronic kidney disease and liver disease were not likely differential diagnoses for the anemia, as none of their functional parameters were abnormal at the time. The cat did not have any history of toxin ingestion, making this less likely. It could be argued that, if these pathologic processes were the underlying reasons for agranulocytosis and anemia, then the cell counts would not have been improved solely by discontinuing methimazole.

Another limitation of this case report was the inability to make a connection between the development of hepatotoxicity and methimazole. The elevations in ALT, ALP, and total bilirubin were consistent with the hepatotoxic findings reported by Peterson et al (1). However, this cat also had an elevation in GGT not reported in this article that suggests a process affecting the biliary epithelium, such as cholangiohepatitis (9). The elevated globulins and neutrophils also seen at the D10 visit suggested an inflammatory process such as cholangiohepatitis (9). However, the changes in liver markers and icterus occurred 9 d after methimazole treatment was stopped, and the icterus only subjectively improved by D20. Peterson et al reported that changes consistent with hepatotoxicity resolved within 7 to 45 d after discontinuing methimazole (1). Thus, since the hepatotoxicity did not develop until after methimazole was stopped and because no bloodwork was done on D20 to assess if the hepatotoxic markers improved, it is not certain whether these changes were due to methimazole.

This case report affirms that methimazole can result in the development of agranulocytosis. Furthermore, it adds another case to the literature on anemia linked to methimazole administration. In effect, it is strongly recommended that RBC counts be regularly checked during the first 3 mo after beginning methimazole treatment, as the development of a severe anemia can be fatal. In addition, future studies should seek to evaluate the exact mechanism through which antithyroid drugs cause anemia.