Beginning in early 2007, reports of renal failure in cats and dogs fed various commercial pet foods led to concerns about possible contamination, and a widespread recall of suspect foods was initiated prior to identification of the toxin. Subsequent investigation revealed melamine and cyanuric acid both in the suspect foods and in the imported wheat gluten used in the manufacture of those foods. This discovery led to further product recalls, resulting in the largest cat and dog food recall in US history. Various reports linked the development of renal failure in cats and dogs with feeding of the recalled food; however, such reports were often anecdotal, and findings may have been complicated by preexisting or concurrent disease or by differences in husbandry or environment. This study documents clinicopathologic, histologic, and toxicologic findings in a well-defined population of 70 cats that were inadvertently exposed to dietary melamine and cyanuric acid in commercially prepared wet-food canned or pouched diets.
Materials and Methods
Animals—Seventy cats from a single, well-maintained commercial cattery were included in the study. Cats ranged in age from 2 to 17 years and ranged in weight from 2.5 to 6.8 kg (5.5 to 15 lb). Cats had been kept at the cattery for at least 9 months, where they were used for routine palatability and acceptability testing of commercial cat foods. They were individually housed in elevated plastic runs inside rooms holding 80 to 100 cats; water was provided ad libitum in stainless steel bowls.
Cats were inadvertently exposed to contaminated commercial canned or pouched food during dietary trials. Four groups of cats exposed to different contaminated commercial diets were identified. Cats in groups 1 (20 cats) and 2 (20 cats) were involved in trials that compared consumption of 2 diets fed simultaneously to determine diet preference. Cats in groups 3 (20 cats) and 4 (10 cats) were involved in trials during which a single diet was offered to determine diet acceptability. Access to the contaminated diets varied. Cats in group 1 were offered the contaminated food for 6 days (2 h/d), cats in group 2 were offered the contaminated food for 4 days (2 h/d), and cats in groups 3 and 4 were offered the contaminated food for 4 days (22 h/d). Dietary trials for cats in groups 1, 2, and 3 were completed before the recall was announced. Cats in group 4 were fed a diet that was not on the original recall list approximately 1 week later. The dietary trial for cats in group 4 was prematurely ended when animals became lethargic and anorexic. The amount of the contaminated diet that was consumed was not quantified for cats in groups 1 and 2. For cats in groups 3 and 4, the amount of the contaminated diet that was consumed was quantified by subtracting the amount of food remaining after each 22-hour period from the amount of food offered. All 4 diets met the Association of American Feed Control Officials standards for labeling as providing complete and balanced nutrition.
Clinical signs—Clinical signs were recorded for all cats that developed any abnormalities after eating the contaminated food. Cats with anorexia, dehydration, and vomiting were treated with amoxicillin (20 mg/kg [9.1 mg/lb], PO, q 12 h) and marbofloxacin (2.75 mg/kg [1.25 mg/lb], PO, q 24 h). Cats with clinical evidence of dehydration were treated with SC administration of a balanced electrolyte solution as needed.
Clinicopathologic findings—Blood samples were obtained from the 30 cats in groups 3 and 4 prior to the start of the dietary trials and submitted for a CBC and serum biochemical analyses. Blood samples were obtained from 68 available cats (1 group 1 cat and 1 group 2 cat died prior to blood collection) between 7 and 11 days after initial exposure to the contaminated food and submitted for serum biochemical analyses. Samples from 67 cats were submitted for CBCs. Additional blood samples were obtained as necessary to assess progression of disease and recovery in affected animals. All samples were submitted to a commercial laboratorya for analysis.
Cats that were azotemic with urine specific gravity < 1.035 were considered to have renal failure. For cats in groups 1 and 2, renal failure was classified as acute if the onset of clinical signs was abrupt and the animal recovered. For cats in groups 3 and 4, renal failure was classified as acute if results of clinicopathologic testing prior to the dietary trial were within reference ranges. For animals that died, necropsy results were used to confirm the diagnosis of acute renal failure. For animals that survived, renal function was considered to have returned to normal if SUN and serum creatinine concentrations returned to reference ranges. Progression to chronic renal failure was diagnosed if azotemia did not resolve.
In selected instances, urine was obtained from affected cats by means of cystocentesis or free catch and submitted for analysis. In addition, urine samples were obtained from all cats that underwent a necropsy.
Pathologic findings—For cats that died or were euthanized, a necropsy was performed at the commercial cattery or at the University of Pennsylvania School of Veterinary Medicine. Representative tissue samples were submitted to the University of Pennsylvania for histologic examination. Tissue samples were routinely processed, and sections (5 μm thick) were stained with H&E.
For kidney sections, severity of tubular necrosis was classified as mild (< 25% of the kidney affected), moderate (≥ 25% but < 75% of the kidney affected), or severe (≥ 75% of the kidney affected). Severity of crystalluria was assessed by examining 10 fields chosen at random at 100× magnification under polarized light. Crystalluria was classified as mild if < 25 tubules contained crystals, as moderate if ≥ 25 but < 50 tubules contained crystals, and as severe if ≥ 50 tubules contained crystals. Crystal morphology was recorded. Finally, the degree of perivascular inflammation was considered to be mild if vessels were surrounded by lymphocytes, plasma cells, or histiocytes; moderate if fibrin with aggregates of neutrophils or eosinophils and a few fibrin thrombi were also seen; and severe if numerous (ie, > 3) fibrin thrombi and large aggregates (ie, > 20 cells) of neutrophils or eosinophils were present.
Toxicology—Samples of the diets were tested for various mycotoxins (ie, aflatoxin B1, aflatoxin B2 aflatoxin G1, aflatoxin G2, diacetoxyscripinol, deoxynivalenol, sterigmatocystin, ochratoxin, T-2 toxin, and zearalenone) by means of thin-layer chromatography following use of a modified Romer procedure1 for sample extraction. In brief, samples were extracted with a 40:60 mixture of methanol and water, followed by liquid-liquid partitioning and cleanup. Sample extracts were then concentrated and spotted, along with standards, on thin-layer chromatography plates. Plates were developed by use of a 2-dimensional solvent procedure. The detection limit was 2 to 5 ppb for the various aflatoxins and 20 to 50 ppb for the other mycotoxins.
Thin-layer chromatography was also used to screen urine samples for sulfonamide residues. The urine was extracted under acidic conditions by means of liquid-liquid extraction. Extracts were concentrated under a stream of nitrogen in a 60°C water bath and then spotted on thin-layer chromatography plates. Plates were developed in chloroform:ethanol (9:1) and cyclohexane: chloroform:acetic acid (4:4:2) solvent systems. Developed plates were dried, sprayed with fluorescamine, and heated. Plates were then viewed under short-wave UV light (254 nm) and long-wave UV light (274 nm) for the bright green-yellow fluorescence characteristic of sulfonamides. Plates were then sprayed with modified Ehrlich solution, heated, and examined for the magenta color characteristic of sulfonamides.
Liquid chromatography–tandem mass spectrometry performed in the selected reaction monitoring mode as described2 was used to test samples of the diets for the ionophore antimicrobials lasalocid and monensin. Extracts from a portion of homogenized food were prepared by means of liquid-solid and liquid-liquid extraction techniques. Food extracts were further purified by a defatting step and a solvent wash step and then concentrated to dryness. Food extract residues were reconstituted in the liquid chromatography mobile phase (1 mL), and an aliquot (5 μL) of each sample was injected into the liquid chromatography–tandem mass spectrometry system.b The system incorporated a C18 (100× 2.0 mm) column coupled to a tandem triple quadrupole mass spectrometer equipped with a liquid chromatography–tandem mass spectrometry interface. The limit of detection was 0.04 ppb.
Gas chromatography–mass spectometry was used to perform qualitative analyses of tissue, urine, and food samples for drugs and pesticides following solid-liquid extraction with pH modification for acidic and basic compounds. Food and tissue samples were also analyzed for melamine and cyanuric acid following solid-liquid extraction. Briefly, pureed samples (4 g) were extracted with a 50:50 mixture of methanol and water (20 mL), which was acidified by addition of 0.6 N hydrochloric acid (0.6 mL). After a 10-minute extraction time, samples were centrifuged and 5 mL of the supernatant was transferred to a clean tube. Samples were then extracted by means of liquid-liquid extraction under basic (pH, 11.0) conditions with ethyl acetate. Urine samples were extracted by means of liquid-liquid extraction under basic (pH, 11.0) conditions with ethyl acetate. The ethyl acetate solvent layer was evaporated to dryness under a stream of nitrogen gas in a 60°C water bath. Food, tissue, and urine extracts were then derivatized with acetonitrile (100 μL) and BSTFA with 1% TMCS (100 μL) at 100°C for 1 hour. Derivatized extracts (1 μL) were injected by means of splitless injection into the mass spectrometer.c The mass spectrometer was operated in full scan mode with a mass range of 40 to 550 amu and electron impact ionization at 70 eV.
Food, liver, and kidney samples were tested for trace mineral content (calcium, phosphorus, magnesium, sodium, potassium, iron, manganese, zinc, and molybdenum) by means of a modification of an established method.3 Samples were weighed, dried in an oven, ashed at 500°C, and dissolved in nitric acid and hydrochloric acid. Digests were centrifuged and analyzed by means of radial and axial inductively coupled argon plasma atomic emission spectroscopy.d
Additional food, liver, and kidney samples were digested routinely in a microwave ovene prior to analysis for heavy metal content (arsenic, cadmium, lead, selenium, and mercury) as described.f Samples were weighed, dried in an oven, and digested in polytef vessels with concentrated nitric acid (7 mL) and, subsequently, 30% hydrogen peroxide (3 mL). For determination of arsenic, cadmium, and lead content, digests were centrifuged and analyzed by means of inductively coupled argon plasma atomic emission spectroscopy. For determination of selenium and mercury content, digests were further treated with 5% sulfamic acid and 50% hydrochloric acid prior to analysis as described.4 The digest was pumped to a hydride generator and mixed with 0.5% sodium borohydride in 0.1% sodium hydroxide, releasing selenium hydride and elemental mercury, which was carried by argon flow to the axial plasma torch for analysis.
Statistical analysis—The Wilcoxon rank sum test was used to compare age between cats that survived and cats that died or were euthanized.
Results
Clinical findings—For cats in groups 3 and 4, the mean quantity of test diet consumed by each cat was 272 g (range, 87 to 475 g) on day 1, 110 g (range, 5 to 302 g) on day 2, 69 g (range 12 to 270 g) on day 3, and 58 g (range, 11 to 216 g) on day 4.
Forty-three of 70 cats had clinical signs; clinical signs varied in severity. Inappetence with or without vomiting was the only sign in 8 cats. Thirty-five cats developed polydipsia, polyuria, dehydration, vomiting, lethargy, and anorexia. Depression, weakness, and dyspnea were observed rarely. Abdominal palpation revealed grossly enlarged kidneys in 7 cats and a greatly distended urinary bladder in 1 cat.
Clinicopathologic findings—One cat in group 1 was euthanized prior to submission of blood samples for clinicopathologic testing, and 1 other cat from this group died before a blood sample could be obtained for a CBC. Fifteen of the remaining 19 cats in group 1 were azotemic when tested 7 to 11 days after consuming the contaminated food, including 12 cats with SUN concentration > 100 mg/dL and 6 cats with serum creatinine concentration > 8 mg/dL. Urine specific gravity was < 1.035 in 5 of 6 cats in which it was measured. One of these 15 cats died and 6 were euthanized. Five weeks after consumption of the contaminated food, 5 of the azotemic cats had recovered renal function and 3 were still azotemic. However, SUN concentration was < 74 mg/dL and serum creatinine concentration was < 3.8 mg/dL in all 3 azotemic cats. Seven weeks after consumption of the contaminated food, 1 of the 3 azotemic cats had recovered renal function, 1 was euthanized because of chronic renal failure (SUN concentration, 65 mg/dL; serum creatinine concentration, 3.1 mg/dL) and chronic nonregenerative anemia, and 1 was still azotemic (SUN concentration, 51 mg/dL; serum creatinine concentration, 2.5 mg/dL). The latter cat recovered renal function 22 weeks after consumption.
One cat in group 2 was euthanatized prior to submission of blood samples for clinicopathologic testing, and 2 of the remaining 19 cats in group 2 were azotemic when tested 7 to 11 days after consuming the contaminated food. One of these cats had an SUN concentration of 177 mg/dL and a serum creatinine concentration of 7.3 mg/dL and was euthanatized. The other cat had improved renal function 7 weeks after consumption of the contaminated food, and SUN and serum creatinine concentrations were similar to values obtained the previous year.
Results of pre-exposure clinicopathologic testing were normal in all group 3 cats. Thirteen of the 20 cats in group 3 were azotemic when tested 9 days after consuming the contaminated food, including 5 cats with SUN concentration > 100 mg/dL and 1 cat with serum creatinine concentration > 8 mg/dL. Urine specific gravity was < 1.035 in 1 of 7 urine samples tested. One of the 13 azotemic cats was euthanatized; the other 12 cats had recovered renal function 4 weeks later.
Results of pre-exposure clinicopathologic testing were normal in all group 4 cats. Eight of the 10 cats in group 4 were azotemic when tested 7 days after consuming the contaminated food, including 2 cats with SUN concentration > 100 mg/dL and serum creatinine concentration > 8 mg/dL. Urine specific gravity was < 1.035 in 3 of 5 urine samples tested. One of the 8 azotemic cats was euthanatized. The other 7 cats had recovered renal function by 4 weeks after consuming the contaminated food, although 1 of these cats was euthanatized because of nonregenerative anemia and lethargy.
In total, azotemia was documented in 38 of the 68 cats for which serum biochemical analyses were performed 7 to 11 days after consumption of the contaminated food (Table 1). Cats with severe azotemia, hyperphosphatemia (16.4 to 31.6 mg/dL), and hyperkalemia (6.2 to 9.4 mEq/ L) did not survive. Urine samples from 18 cats with azotemia were submitted for analysis, and urine specific gravity was < 1.035 in 9. Microscopic examination of urine sediment from 8 cats in group 1 revealed the presence of green-brown circular crystals (Figure 1).
Results of serum biochemical analyses performed in 68 cats 7 to 11 days after inadvertent exposure to pet food contaminated with meiamine and cyanuric acid.
| Variable | Cats with abnormal values | Reference range9 | |
|---|---|---|---|
| No. | Range | ||
| Urea nitrogen (mg/dL) | 38 | 41–415 | 14–36 |
| Creatinine (mg/dL) | 32 | 2.5–28.6 | 0.6–2.4 |
| Amylase (U/L) | 28 | 1,205–2,203 | 199–1,200 |
| Phosphorus (mg/dL) | 17 | 8.3–30.9 | 2.4–8.2 |
| Potassium (mEq/L) | 3 | 6.2–9.4 | 3.4–5.6 |
Photomicrographs of circular green-brown crystals in urine from a cat that ingested food contaminated with melamine and cyanuric acid (A) and irregular needle-shaped crystals obtained by adding melamine to urine from a cat (B).
Citation: Journal of the American Veterinary Medical Association 233, 5; 10.2460/javma.233.5.729
Photomicrographs of circular green-brown crystals in urine from a cat that ingested food contaminated with melamine and cyanuric acid (A) and irregular needle-shaped crystals obtained by adding melamine to urine from a cat (B).
Citation: Journal of the American Veterinary Medical Association 233, 5; 10.2460/javma.233.5.729
Photomicrographs of circular green-brown crystals in urine from a cat that ingested food contaminated with melamine and cyanuric acid (A) and irregular needle-shaped crystals obtained by adding melamine to urine from a cat (B).
Citation: Journal of the American Veterinary Medical Association 233, 5; 10.2460/javma.233.5.729
Results of a CBC performed 7 to 11 days after consumption of contaminated food were available for 67 cats, and clinical anemia (PCV < 29) was documented in 8, 5 of which were euthanatized at that time. Nonregenerative anemia was documented 3 weeks after exposure in the 3 surviving anemic cats and in an additional 4 previously nonanemic, azotemic cats. The anemia resolved in 2 cats by week 7 and 1 cat by week 22; 2 anemic cats were euthanatized, and 2 were still anemic 22 weeks after consumption of the contaminated food.
Mean age of the 56 cats that survived, including clinically unaffected animals (mean ± SD, 6.3 ± 3.4 years; range, 2 to 12 years), was significantly lower than mean age of the 14 cats that died or were euthanatized (10.3 ± 3.4 years; range, 3 to 17 years; P < 0.001). Only 1 cat that was euthanatized was < 8 years old.
Pathologic findings—Overall, 1 cat died, and 13 were euthanized. Twelve cats were necropsied at the commercial cattery, and 2 were submitted to the University of Pennsylvania for necropsy. Tissue samples from 13 of the 14 cats were submitted for histologic evaluation. In general, renal lesions did not vary across groups. Therefore, results for all cats that died or were euthanized were reported together.
On gross examination, 10 of the 14 cats had bilateral enlargement of the kidneys. In 1 cat, the right kidney was small, nodular, and firm, whereas the left kidney was large and swollen, and in 2 cats, both kidneys were small and nodular. The remaining cat had normal-sized kidneys with a focal depression in the left renal cortex. Other gross abnormalities included bilateral white plaques on the ventral surface of the tongue (1 cat) and multifocal white plaques on the gastric mucosa with associated hyperemia (5 cats). Two cats had ascites, which was mild in 1 cat and severe in the other. In the cat with severe ascites, the liver and pancreas were both nodular and firm. Two cats had large cystic hepatic masses, 1 of which had an adherent blood clot. The caudal lung lobes of 1 cat had multifocal, firm, nodular, and plaque-like masses involving the pleura and much of the parenchyma. One cat had moderate gingivitis, and another had multiple blue-black masses within the ear canals.
Kidneys of 13 cats were examined histologically, and all had aggregates of gold-brown crystals within the distal segments of the nephron, classified as mild (1 cat), moderate (9 cats), or severe (3 cats). Most crystals were large (15 to 30 μm in diameter) and consisted of 2 distinct concentric rings with linear striations that gave the appearance of spokes radiating from the center (Figure 2). Fractured crystals of this type were semicircular to wedge-shaped. Other crystals consisted of small (2 to 3 μm in diameter), single-to-multiple refractile granules (Figure 3). Additional clear crystals with a sheaves-of-wheat morphology and irregular, variably sized shards were occasionally observed but were less common. Crystals were frequently associated with sloughed epithelial cells.
Photomicrograph of a section of kidney from a cat euthanatized approximately 2 weeks after eating pet food contaminated with melamine and cyanuric acid. Notice the large (20 to 30 μm in diameter) gold-brown circular crystals (arrow) in a renal tubule lined by regenerating, attenuated epithelium. Crystals consisted of 2 distinct concentric rings with linear striations that gave the appearance of spokes radiating from the center. H&E stain; bar = 50 μm.
Citation: Journal of the American Veterinary Medical Association 233, 5; 10.2460/javma.233.5.729
Photomicrograph of a section of kidney from a cat euthanatized approximately 2 weeks after eating pet food contaminated with melamine and cyanuric acid. Notice the large (20 to 30 μm in diameter) gold-brown circular crystals (arrow) in a renal tubule lined by regenerating, attenuated epithelium. Crystals consisted of 2 distinct concentric rings with linear striations that gave the appearance of spokes radiating from the center. H&E stain; bar = 50 μm.
Citation: Journal of the American Veterinary Medical Association 233, 5; 10.2460/javma.233.5.729
Photomicrograph of a section of kidney from a cat euthanatized approximately 2 weeks after eating pet food contaminated with melamine and cyanuric acid. Notice the large (20 to 30 μm in diameter) gold-brown circular crystals (arrow) in a renal tubule lined by regenerating, attenuated epithelium. Crystals consisted of 2 distinct concentric rings with linear striations that gave the appearance of spokes radiating from the center. H&E stain; bar = 50 μm.
Citation: Journal of the American Veterinary Medical Association 233, 5; 10.2460/javma.233.5.729
Photomicrograph of a section of kidney from a cat euthanatized approximately 2 weeks after eating pet food contaminated with melamine and cyanuric acid. Notice the cluster of small (2 to 3 μm in diameter) gold-brown granules (arrow) with admixed neutrophils inside a tubule lined by regenerative epithelium. H&E stain; bar = 20 μm.
Citation: Journal of the American Veterinary Medical Association 233, 5; 10.2460/javma.233.5.729
Photomicrograph of a section of kidney from a cat euthanatized approximately 2 weeks after eating pet food contaminated with melamine and cyanuric acid. Notice the cluster of small (2 to 3 μm in diameter) gold-brown granules (arrow) with admixed neutrophils inside a tubule lined by regenerative epithelium. H&E stain; bar = 20 μm.
Citation: Journal of the American Veterinary Medical Association 233, 5; 10.2460/javma.233.5.729
Photomicrograph of a section of kidney from a cat euthanatized approximately 2 weeks after eating pet food contaminated with melamine and cyanuric acid. Notice the cluster of small (2 to 3 μm in diameter) gold-brown granules (arrow) with admixed neutrophils inside a tubule lined by regenerative epithelium. H&E stain; bar = 20 μm.
Citation: Journal of the American Veterinary Medical Association 233, 5; 10.2460/javma.233.5.729
Renal tubular necrosis was observed in all 13 cats and was classified as severe in 3 cats, moderate in 9, and mild in 1. Tubular necrosis was characterized by dilated tubules lined by epithelial cells with hypereosinophilic cytoplasm and pyknotic or karyorrhectic nuclei. Sloughed necrotic cells were often present within tubular lumina. Tubular rupture was a consistent feature. In some cats, tubules were lined by attenuated epithelial cells or by closely clumped cuboidal cells with basophilic cytoplasm and high nuclear-to-cytoplasmic ratios, indicative of tubular regeneration. Similar tubular lesions were seen in the small, nodular kidneys, with the addition of interstitial fibrosis and multifocal infiltrates of lymphocytes and plasma cells. Within these foci, glomeruli were often sclerotic and collapsed.
Perivascular inflammation involving the renal subcapsular veins was identified in 12 of the 13 cats (Figure 4). Two cats with moderate to severe perivascular inflammation also had involvement of an interlobular vein. Lesions were characterized by disruption of vessel walls by numerous neutrophils, eosinophils, and macrophages. In addition, affected vessels were surrounded and infiltrated by plump fibroblasts. Endothelial cells were hypertrophied and fibrin thrombi were present in 6 cats. Perivascular inflammation was classified as severe in 2 cats, moderate in 7, and mild in 3.
Photomicrograph of a section of kidney from a cat euthanatized approximately 2 weeks after having been inadvertently fed pet food contaminated with melamine and cyanuric acid. Notice that the wall of the subcapsular vein is disrupted by neutrophils, macrophages, eosinophils, and lymphocytes. A small fibrin thrombus (arrow) is adherent to the tunica intima. H&E stain; bar = 100 μm.
Citation: Journal of the American Veterinary Medical Association 233, 5; 10.2460/javma.233.5.729
Photomicrograph of a section of kidney from a cat euthanatized approximately 2 weeks after having been inadvertently fed pet food contaminated with melamine and cyanuric acid. Notice that the wall of the subcapsular vein is disrupted by neutrophils, macrophages, eosinophils, and lymphocytes. A small fibrin thrombus (arrow) is adherent to the tunica intima. H&E stain; bar = 100 μm.
Citation: Journal of the American Veterinary Medical Association 233, 5; 10.2460/javma.233.5.729
Photomicrograph of a section of kidney from a cat euthanatized approximately 2 weeks after having been inadvertently fed pet food contaminated with melamine and cyanuric acid. Notice that the wall of the subcapsular vein is disrupted by neutrophils, macrophages, eosinophils, and lymphocytes. A small fibrin thrombus (arrow) is adherent to the tunica intima. H&E stain; bar = 100 μm.
Citation: Journal of the American Veterinary Medical Association 233, 5; 10.2460/javma.233.5.729
The cat euthanatized 4 weeks after consumption of the contaminated food, which was not azotemic at the time of euthanasia, had moderate perivascular fibrosis. There were multifocal cortical infiltrates of lymphocytes, plasma cells, neutrophils, and eosinophils. Occasional tubules contained necrotic epithelial cells and rare circular crystals. The left kidney had a large focus of papillary tubular mineralization with an adjacent radiating band of fibrosis typical of a chronic infarct.
Kidneys of the azotemic cat euthanatized 8 weeks after consumption of the contaminated food had moderate tubular necrosis and regeneration with foci of dysplasia in the regenerating tubules. Dysplasia was characterized by anisokaryosis, frequent karyomegaly, anisocytosis, and multiple large nucleoli. Intratubular crystals were abundant and often mineralized. Subcapsular venous lesions were scored as severe because all were severely stenotic and surrounded by dense fibrosis (Figure 5).
Photomicrograph of a section of kidney from a cat euthanatized approximately 8 weeks after inadvertent exposure to pet food contaminated with melamine and cyanuric acid. Notice that the subcapsular vein (arrow) is stenotic and surrounded by dense fibrosis. Masson trichrome stain; bar = 50 μm.
Citation: Journal of the American Veterinary Medical Association 233, 5; 10.2460/javma.233.5.729
Photomicrograph of a section of kidney from a cat euthanatized approximately 8 weeks after inadvertent exposure to pet food contaminated with melamine and cyanuric acid. Notice that the subcapsular vein (arrow) is stenotic and surrounded by dense fibrosis. Masson trichrome stain; bar = 50 μm.
Citation: Journal of the American Veterinary Medical Association 233, 5; 10.2460/javma.233.5.729
Photomicrograph of a section of kidney from a cat euthanatized approximately 8 weeks after inadvertent exposure to pet food contaminated with melamine and cyanuric acid. Notice that the subcapsular vein (arrow) is stenotic and surrounded by dense fibrosis. Masson trichrome stain; bar = 50 μm.
Citation: Journal of the American Veterinary Medical Association 233, 5; 10.2460/javma.233.5.729
Hepatic sinusoidal dilatation (telangiectasia) was observed in 2 cats. Biliary hyperplasia with associated portal fibrosis was seen in 3 cats and was classified as mild in 2 cats and moderate in 1. Biliary cystadenoma was diagnosed in one of the cats with a hepatic mass. The other hepatic mass was a large (3 cm in diameter) hyperplastic nodule. Within the nodule, hepatocytes were disorganized, and there were a few randomly scattered portal areas. Atypical vascular channels were present within the mass and often contained fibrin thrombi. A large organizing hematoma was adhered to the surface of the mass. The cat with the firm nodular liver had severe bridging portal fibrosis with bile duct proliferation and multifocal chronic inflammation.
Severe chronic fibrosing lymphocytic pancreatitis with pancreatic nodular hyperplasia was observed in the cat with bridging portal fibrosis. One cat had mild, multifocal, chronic, lymphocytic, interstitial pancreatitis.
Gastric uremic mineralization was present in 3 cats. Three cats had mild to moderate multifocal necrosis of the gastric glandular epithelial cells with associated large spiral bacteria within parietal cells and gastric pits. Additionally, a cross section of a single nematode parasite was present within the gastric glands of 1 of these 3 cats. The intestine of 1 cat contained a mild increase in the number of lymphocytes, plasma cells, and neutrophils in the lamina propria with associated fibrosis.
The heart of 1 cat had marked endomyocardial fibrosis with subjacent myocardial necrosis and neutrophilic and histiocytic secondary inflammation.
Mild myeloid hyperplasia was present in 1 cat with anemia. Evaluation of the bone marrow in another cat with anemia did not reveal any abnormalities. The spleen of 1 cat had perivascular accumulations of macrophages (ellipsoids). Severe chronic interstitial pneumonia with fibrosis and type II pneumocyte hyperplasia and dysplasia was present in the cat with firm lungs. For the remaining organs, results of histologic examination were unremarkable.
Toxicologic findings—Melamine and cyanuric acid were detected by means of gas chromatography–mass spectometry in diets fed to group 1 cats and were detected in 3 samples of wheat gluten used in the production of all of the diets. Quantitative analyses were not performed.
Results of thin-layer chromatography for mycotoxins in the diets were negative, as were results of liquid chromatography–tandem mass spectrometry for the ionophore antibiotics monensin and lasalocid. No glycols, such as ethylene glycol, diethylene glycol, propylene glycol, and triethylene glycol, or clinically important drugs or pesticides were detected in the diets by means of gas chromatography–mass spectometry.
Gas chromatography–mass spectometry revealed melamine in all 12 urine samples submitted for toxicologic analysis; 6 of 8 urine samples tested were positive for cyanuric acid. Melamine and cyanuric acid were detected in 2 kidney samples and 1 vomitus sample submitted for qualitative analysis but were not detected in liver samples. Sulfonamides were not detected in any of the 4 kidney and urine samples tested by means of thin-layer chromatography. One liver sample contained pentobarbital, and the kidney and liver samples from another cat contained pentobarbital and phenytoin.
Analysis yielded 1.45% to 1.51% calcium in 3 samples of diets fed to group 1 cats, 1.37% to 1.43% phosphorus, 18.53 to 18.81 mg/kg copper, and 1.18 to 1.29 mg/kg selenium. The iron concentration ranged from 4,804.97 to 4,887.47 mg/kg. Lead, arsenic, cadmium, and mercury were present in trace concentrations (0.48 to 0.55 mg/kg, 0.026 to 0.031 mg/kg, 0.051 to 0.052 mg/kg, and 0.005 to 0.009 mg/kg, respectively). Diets fed to cats in the other 3 groups were not submitted for trace mineral and heavy metal analyses. Kidney and liver concentrations of zinc were within reference ranges.5 Hepatic iron concentrations were adequate to high, ranging from 797.3 to 1,694.3 mg/kg (reference range, 279.0 to 1,000.0 mg/kg), but renal iron concentrations were within reference limits.5 Hepatic copper concentrations were less than the lower reference limit (57.2 to 101.8 mg/kg; reference range, 129.5 to 180.0 mg/kg), and kidney copper concentrations were normal or low (6.7 to 9.7 mg/ kg; reference range, 7.7 to 10.8 mg/kg).5 Liver calcium concentrations were within reference limits. All renal calcium concentrations were high (921.3 to 1,694.5 mg/ kg; reference range, 164.5 to 400.0 mg/kg).5 All values are given on a dry matter basis.
Discussion
In the present study, 43 of 70 cats developed signs of toxicosis following inadvertent exposure to pet food contaminated with melamine and cyanuric acid. The most consistent clinical and pathologic abnormalities were associated with the urinary tract, and histologic examination revealed tubular necrosis and crystalluria. Toxicologic analysis verified the presence of melamine and cyanuric acid in samples of the pet foods and in kidney and urine samples from affected cats.
The severity of clinical, clinicopathologic, and histopathologic signs of toxicosis varied among and within groups in the present study. One potential reason for this variation was that each of the 4 groups was fed a different diet. Furthermore, within each group, cats were exposed to cans and pouches from multiple lots. Concentrations of melamine and cyanuric acid likely varied within and among diet types and lotsh; therefore, the amount of toxin to which cats were exposed also varied. In addition, our findings suggest that the susceptibility to the toxic effects of melamine and cyanuric acid may have varied with age because cats that died were significantly older than cats that survived. It is possible that older cats had subclinical renal disease that was exacerbated by toxin exposure. Finally, individual differences in urine pH and food intake also played a role in the variation in severity of signs of toxicosis.
Thirty-eight of 68 (56%) cats were azotemic 7 to 11 days after exposure to the contaminated food, and urine specific gravity was < 1.035 in 9 of 18 cats evaluated. Acute renal failure was diagnosed in 16 cats, including 13 that died or were euthanatized, 1 cat from group 3, and 2 cats from group 4 that recovered normal renal function. Clinicopathologic findings were consistent with acute renal failure in the remaining 22 cats, although urine was not obtained for confirmation. One cat progressed to chronic renal failure and was euthanatized at 8 weeks, but the surviving cats eventually recovered or had an improvement in renal function.
Histologic identification of renal tubular necrosis with regeneration is consistent with exposure to a nephrotoxin. In mice, rats, and rabbits, tubular regeneration begins within 3 days of toxin exposure, although regenerating tubules are morphologically and functionally abnormal until 7 to 14 days after exposure.6 This time frame is consistent with findings in the present study. One cat initially had severe azotemia with subsequent recovery of renal function after 4 weeks, and histologic examination of the kidneys revealed only mild tubular necrosis. Many tubules were regenerating, but most were normal, suggesting that tubular regeneration corresponded with clinical recovery from renal failure.
Previously reported causes of nephrotoxicosis in cats include ethylene glycol, sulfonamides, and lily plants.7 Results of cytologic examination of urine samples and histologic examination of kidney specimens from cats in the present study were not consistent with ethylene glycol toxicosis. Cats with ethylene glycol toxicosis have numerous intratubular calcium oxalate crystals, consisting of irregular shards or clusters with a sheaves-of-wheat morphology.8 The distinct gold-brown circular crystals seen in cats in the current study have not been reported with ethylene glycol toxicosis. The infrequent irregular shards and sheaves of wheat identified in the present study were attributed to the inability of the diseased kidneys to process endogenous oxalates.9 Furthermore, neither ethylene glycol nor related compounds were detected in food or tissue samples. Finally, renal calcium concentrations in cats with ethylene glycol toxicosis are typically > 30,000 mg/kg on a dry matter basis,10 whereas concentrations for cats in the present study were < 1,700 mg/kg.
Toxins that reportedly can cause renal tubular necrosis without associated crystalluria in cats include vitamin D (cholecalciferol), NSAIDs, mycotoxins, diethylene glycol, and heavy metals.11–13 Histologic lesions in cats in the present study were not consistent with those expected with cholecalciferol toxicosis. Nonsteroidal anti-inflammatory drugs, mycotoxins, and diethylene glycol were not detected in food samples. Although trace concentrations of heavy metals were detected in the food, the concentrations were below tolerance limits established by the National Research Council.14
Perivascular inflammation involving the renal subcapsular veins was identified in 12 of 13 cats in the present study. The pathogenesis of this lesion is not currently known, although similar lesions have been observed in cats with feline infectious peritonitis6 and in renal transplant recipients.15 One possible cause is damage to the subcapsular veins due to stretching of the vessel, which would result from edema-induced renomegaly. Any disruption in the integrity of the endothelial basement membrane allows fibrin to adhere to the subendothelial collagen and enables inflammatory cells to escape the vessel lumen.6 In cats, the renal subcapsular veins are large and drain the outer renal cortex.16 They travel along the exterior surface of the kidney and empty directly into the renal vein, whereas in dogs, these veins travel through the cortex to communicate with the arcuate and interlobular veins.16 This specialized renal vascular anatomy may make cats less tolerant of acute renomegaly. It is also possible that the subcapsular perivascular inflammation was related to the inflammatory response incited by the tubular necrosis and tubulorrhexis. As inflammatory cells exit through the venules, they become activated, produce cytokines, and release noxious agents, such as reactive oxygen species6; therefore, perivascular inflammation may represent a bystander lesion.
The finding of severe perivenous fibrosis and venous luminal stenosis in the cat euthanatized 8 weeks after exposure is intriguing. Unfortunately, diagnostic imaging (eg, contrast-enhanced ultrasonography) to assess the functional capabilities of these vessels was not performed. Nevertheless, these findings suggest that cats that survive acute renal failure may be predisposed to developing chronic renal disease secondary to ischemia.
Lesions secondary to uremia (gastric uremic mineralization) were present in 2 cats. Bilateral white plaques on the ventral surface of the tongue were observed in 1 cat with gastric mineralization. Although this lesion was not examined histologically, the appearance and location were considered pathognomonic for uremic glossitis.6
Organs other than the kidney were generally unaffected in cats in the present study. Severe bridging portal fibrosis with bile duct hyperplasia was seen in 1 cat; however, the chronicity of this lesion suggested that it was a preexisting lesion unrelated to the acute intoxication. The large, hyperplastic, hepatic nodule in 1 cat also most likely developed prior to exposure. The vascular proliferation within the nodule was bizarre, and it was difficult to determine when these vessels formed. Hepatocellular proliferations (ie, nodular hyperplasia, adenoma, and carcinoma) are documented to have atypical vascular channels.17 Liver lesions in the other cats (biliary hyperplasia and telangiectasia) were mild and nonspecific and are common incidental findings in older cats.6
Gastric glandular necrosis was observed in 3 cats in the present study. The presence of large spiral bacteria in all 3 cats and the intraglandular nematode in 1 suggested an infectious cause.
Endomyocarditis was observed in 1 cat. Although cardiotoxicosis could not be ruled out, the necrotic foci were consistently associated with inflammation and were interpreted to be secondary.
Nonregenerative anemia was diagnosed > 3 weeks after exposure in 7 cats in the present study that had previously been azotemic. Possible causes included a direct toxic effect on the bone marrow or decreased erythropoietin production by the damaged kidneys.18 Although blood loss through the gastrointestinal tract can also cause anemia,18 no ulcers were identified grossly or histologically. Furthermore, gastrointestinal tract blood loss results in regenerative anemia.18 Bone marrow was examined in 3 of the anemic cats, and no evidence of direct toxic insult was observed. Because anemia was present only in cats that were previously azotemic, decreased erythropoietin production may have played a role.
Melamine was detected both in the diets fed to the cats in the present study and in samples of imported wheat gluten used in the manufacture of those diets. It was also detected in kidney and urine samples from affected cats. According to the US FDA, melamine was likely added to increase the nitrogen content of wheat flour or poor-quality wheat gluten.19 Protein content in foodstuffs is usually estimated by measuring total nitrogen content by use of the Kjeldahl method and extrapolating the protein concentration. Addition of a nitrogen source such as melamine would raise the nitrogen and therefore the apparent protein content, thereby increasing the commercial value of the product.19
Melamine is a versatile compound used in the manufacture of plastics, and melamine resins are used in the manufacture of housewares, electrical equipment, laminates, textile dyes, and paper.20 Melamine resins have been added to wheat gluten destined for the textile industry,21 and because of its high nitrogen content, melamine has been proposed for use as a fertilizer.22 Complex dendrimers of melamine have been studied for pharmaceutical use.23
The toxicity of pure melamine is relatively low in mammals. For example, the oral LD50 of melamine is 3,200 mg/kg in male rats and 3,800 mg/kg in female rats.20 Similarly, the oral LD50 is 3,300 mg/kg in male mice and 7,000 mg/kg in female mice.20 However, longterm administration of melamine to laboratory rodents at concentrations ranging from 0.225% to 0.9% of the diet causes urolithiasis.20 Lesions in the urinary bladder, including transitional cell carcinoma, were observed in rats fed diets containing ≥ 0.45% melamine. Neoplasia was associated with the presence of uroliths, and it was theorized that mechanical irritation of the mucosal surface produced epithelial hyperplasia and stimulated tumor formation.20 Renal cortical fibrosis and lymphoplasmacytic nephritis were reported in female rats fed diets containing ≥ 0.45% melamine,20 and urolithiasis was consistently identified in rats fed diets containing 0.3% to 3.0% melamine.24 Melamine-induced crystalluria in sheep has been associated with anuria and uremia secondary to nephrolithiasis.25 Sheep were given single (217 mg/kg) or multiple (1,351 mg/kg/d for 6 days, 510 mg/kg/d for 18 days, or 200 mg/kg/d for 39 days) doses of melamine, and clinical signs, including anorexia and anuria, developed between 5 and 31 days after first exposure in a dose-dependent manner.25 In a study26 involving dogs given melamine (125 mg/kg) as a potential diuretic, no adverse effects were identified, although crystalluria was reported. Crystals were identified as dimelaminemonophosphate and described as fine, white, and needle-shaped crystals (Figure 1),26 unlike the gold-brown circular crystals identified in cats in the present study.
Quantitative analyses were not performed on diets fed to cats in the present study. However, the FDA has tested > 200 pet food samples, with preliminary results showing melamine concentrations ranging from 0 to 2,263 mg/kg.27,h If a cat were to ingest 2 pouches of wet food each day, with each pouch weighing 85 g and containing 0.2% melamine (ie, 2,000 mg/kg), it would be ingesting 340 mg of melamine/d or, for a 4-kg (8.8-lb) cat, 85 mg/kg/d. By comparison, cats in groups 3 and 4 in the present study were eating between 87 and 475 g of food on the first day the contaminated diet was offered.
According to the FDA, various homologs of melamine, such as cyanuric acid, ammeline, and ammelide, were present in recalled diets and may have been generated during melamine production or as a degradation product.28,29 Cyanuric acid was also found in contaminated wheat gluten used in the manufacture of recalled diets, although at lower concentrations than melamine.27 There is little information available concerning the toxicity of these metabolites, although there is basic toxicology information for cyanuric acid.30
Cyanuric acid produces degenerative changes in the kidneys of guinea pigs when administered at a dosage of 30 mg/kg/d for 6 months but not when administered at 10 mg/kg/d.31 Changes have also been identified in rats fed a diet containing 8% monosodium cyanurate for 20 weeks.31 Dogs fed diets containing 8% monosodium cyanurate had lesions similar to those reported in rats, with the additional finding of interstitial fibrosis. Rats and dogs fed diets containing 0.8% monosodium cyanurate did not have any adverse effects.31
An FDA analysis of samples of 4 recalled pet foods revealed melamine at concentrations ranging from 1,573 to 2,236 mg/kg and cyanuric acid at concentrations ranging from 317 to 595 mg/kg.27,h Cats eating 87 to 475 g of food/d, as was the case for cats in groups 3 and 4 in the present study, would ingest approximately 50 to 300 mg of cyanuric acid/d or 12.5 to 75 mg/kg/d for a 4-kg cat. Comparable dietary exposures to cyanuric acid alone did not produce adverse effects or lesions in rats or dogs.31 Renal lesions were produced in guinea pigs given similar doses31 but did not resemble lesions seen in cats in the present study. Acute necrosis of renal tubular epithelium secondary to cyanuric acid ingestion alone has not been documented in any species to the authors' knowledge.
Melamine and cyanuric acid in solution form hydrogen bonds, producing hexamers consisting of 3 melamine and 3 cyanuric acid molecules interlocking in a lattice structure.30 Crystals found in urine of cats and dogs that have eaten food contaminated with melamine and cyanuric acid may represent a similar molecular interaction. A recent study32 found that diets fed to cats containing melamine or cyanuric acid alone at concentrations as high as 1% had no effect on the kidneys. However, cats fed diets containing both melamine and cyanuric acid at concentrations ≥ 0.2% developed clinical signs (vomiting and anorexia) within 12 hours. Renal failure was evident on the basis of high SUN and serum creatinine concentrations and serum electrolyte abnormalities within 36 hours. Cats were euthanatized after 48 hours, and histologic examination revealed renal tubular crystals similar to those found in the present study. Regeneration was not a striking feature, and perivascular lesions were not observed.
Histologic examination of kidney specimens from the cat euthanatized 8 weeks after exposure revealed moderate tubular necrosis with regeneration and crystalluria. Given that the cat had not eaten contaminated food for nearly 8 weeks, the presence of crystals in the kidneys was surprising. The half-life for urinary elimination of melamine in dogs is reported to be 6 hours.26 Thus, it is possible that cyanuric acid altered the elimination kinetics of melamine, perhaps causing crystals to become trapped in the tubules so that they could not be eliminated. It is unlikely that toxins were stored in adipose tissue and released over time because both melamine and cyanuric acid have high polarity and melamine has a low n-octanol:water partition coefficient.28
The combination of melamine and cyanuric acid contamination in pet foods was associated with clinical and clinicopathologic abnormalities of the urinary tract, including azotemia and the presence of unique crystals in the urine. Histologic examination revealed tubular necrosis and similar crystals within tubules. Further evaluation of the survivors will allow assessment of any long-term effects associated with exposure to these 2 toxins.
Antech Diagnostic Laboratories, Southhaven, Miss.
PE-Sciex API 2000, LC/MS/MS system, Perkin-Elmer Corp, Foster City, Calif.
Agilent 5973N GC/MS system, Agilent Technologies, Palo Alto, Calif.
Model DV2100, Perkin-Elmer Instruments, Shelton, Conn.
MARS Xpress, CEM Corp, Matthews, NC.
Modified sample preparation method 5BI-5, CEM Corp, Matthews, NC.
Moroff S, Klaassen J, Antech Diagnostic Laboratory, Lake Success, NY; reference ranges based on testing 40 veterinarianowned, healthy cats from the eastern and western regions of the United States.
Lovell RA, US FDA, Washington, DC: Personal communication, 2007.
References
- 2.↑
Ebel JG, Wachs T, Henion JD. Rapid forensic selected reaction monitoring liquid chromatography/mass spectrometry determination of ionphore antibiotics found at toxic levels in animal feeds. J Assoc Off Anal Chem 2004;87:25–30.
- 3.↑
AOACI. Official method 985.01 metals and other elements in plants and pet foods, inductively coupled plasma spectroscopic method. In: Official methods of analysis of AOAC International. 16th ed. Arlington, Va: AOAC International, 1995;4.
- 4.↑
Davidowski L. A simple continuous flow hydride generator for ICP-OES. In: Perkins-Elmer instruments ICP, application study number: 67. Wilton, Conn: The Perkin-Elmer Corporation, 1993;1–6.
- 5.↑
Puls R. Mineral levels in animal health: diagnostic data. 2nd ed. Aldergrovem, BC, Canada: Sherpa International, 1994;6:313.
- 6.↑
Newman SJ, Confer AW, Panciera RJ. Urinary system. In: McGavin MD, Zachary JF, eds. Pathologic basis of veterinary disease. St Louis: Mosby Inc, 2007;613–692.
- 7.↑
Barsanti JA, Lees GE, Willard MD, et al. Urinary disorders. In: Willard MD, Tvedten H, eds. Small animal clinical diagnosis by laboratory methods. St Louis: Elsevier, 2004;152–153.
- 9.↑
Fogo AB, Kashgarian M. Diagnostic atlas of renal pathology. Philadelphia, Pa: Elsevier Saunders, 2005;393–394.
- 10.↑
Rumbeiha WK. Cholecalciferol. In: Talcott PA, Peterson ME, eds. Small animal toxicology. St Louis: Elsevier, 2006;629–642.
- 11.
DiBartola SP. Clinical approach and laboratory evaluation of renal disease. In: Ettinger SJ, Feldman EC, eds. Textbook of veterinary medicine. Diseases of the dog and cat. 6th ed. St Louis: Elsevier Inc, 2005;1616.
- 12.
Sebastian MM, Baskin SI, Czerwinski SE. Renal toxicity. In: Gupta RC, ed. Veterinary toxicology basic and clinical principles. New York: Elsevier, 2007;167–176.
- 13.
Bischoff K. Diethylene glycol. In: Talcott PA, Peterson ME, eds. Small animal toxicology. St Louis: Elsevier, 2006;693–697.
- 14.↑
National Research Council. Mineral tolerance of animals. 2nd revised ed. Washington, DC: National Academies Press, 2005.
- 15.↑
De Cock HEV, Kyles AE, Griffery SM, et al. Histopathologic findings and classification of feline renal transplants. Vet Pathol 2004;41:244–256.
- 16.↑
Nisson OI, Galskov A. Direct measurement of superficial and deep venous flow in the cat kidney. Circ Res 1972;30:82–96.
- 17.↑
Cullen JM, Popp JA. Tumors of the liver and gall bladder. In: Meuten DJ, ed. Tumors in domestic animals. Ames, Iowa: Iowa State Press, 2002;483–508.
- 18.↑
Weiss D, Tvedten H. Erythrocyte disorders. In: Willard MD, Tvedten H, eds. Small animal clinical diagnosis by laboratory methods. St Louis: Elsevier, 2004;38–62.
- 19.↑
FDA. Press conference on recall of contaminated animal feed. Available at: www.fda.gov/bbs/transcripts/transcript050807.pdf. Accessed Jul 26, 2007.
- 20.↑
Melnick RL, Boorman GA, Haseman JK, et al. Urolithiasis and bladder carcinogenicity of melamine in rodents. Toxicol Appl Pharmacol 1984;72:292–303.
- 21.↑
Scheyer LE, Polsani M. Use of chemically modified wheat gluten to reduce formaldehyde emissions during curing of pigment print pastes on fabrics. Starch 2000;52:420–422.
- 22.↑
Shelton DR, Karns JS. Metabolism of melamine by Klebsiella terragena. Appl Environ Microbiol 1997;63:2832–2835.
- 23.↑
Chen H, Neerman F. Cytotoxicity, hemolysis, and acute in vivo toxicology of dendrimers based on melamine, candidate vehicles for drug delivery. J Am Chem Soc 2004;126:10044–10048.
- 24.↑
Okumura M, Hasegawa R. Relationship between calculus formation and carcinogenesis in the urinary bladder of rats administered the non-genotoxic agents, thymine or melamine. Carcinogenesis 1992;13:1043–1045.
- 26.↑
Lipschitz WL, Stokey E. The mode of action of three new diuretics: melamine, adenine, and formoguanamine. J Pharmacol Exp Ther 1945;82:235–248.
- 27.↑
FDA. Interim melamine and analogues safety/risk assessment. Available at: www.cfsan.fda.gov/∼dms/melamra.html. Accessed Jul 27, 2007.
- 28.↑
Jutzi K, Cook AM. The degradative pathway of the s-traizine melamine. Biochem J 1982;208:679–684.
- 29.
Morales R. Potentiometric titrations of cyanuric acid and melamine in dimethylsulfoxide. Anal Chem 1968;40:1148–1149.
- 30.↑
Xu W, Dong M, Gersen H, et al. Cyanuric acid and melamine on Au(111): structure and energetics of hydrogen-bonded networks. Small 2007;3:854–858.
- 31.↑
Canelli E. Chemical, bacteriological, and toxicological properties of cyanuric acid and chlorinated isocyaurates as applied to swimming pool disinfection, a review. Am J Public Health 1974;64:155–162.
- 32.↑
Puschner B, Poppenga RH. Assessment of melamine and cyanuric acid toxicity in cats. J Vet Diagn Invest 2007;19:616–624.
