Clinical and clinicopathological features and outcomes of cats with suspected dietary induced pancytopenia

Abstract Background After a strong epidemiological link to diet was established in an outbreak of pancytopenia in cats in spring 2021 in the United Kingdom, 3 dry diets were recalled. Concentrations of the hemato‐ and myelotoxic mycotoxins T‐2, HT‐2 and diacetoxyscirpenol (DAS) greater than the European Commission guidance for dry cat foods were detected in the recalled diets. Objectives To describe clinical and clinicopathological findings in cats diagnosed with suspected diet induced pancytopenia. Animals Fifty cats presenting with pancytopenia after exposure to a recalled diet. Methods Multicenter retrospective case series study. Cats with known exposure to 1 of the recalled diets were included if presented with bi‐ or pancytopenia and underwent bone marrow examination. Results Case fatality rate was 78%. Bone marrow aspirates and biopsy examination results were available in 23 cats; 19 cats had a bone marrow aspirate, and 8 cats had a biopsy core, available for examination. Bone marrow hypo to aplasia—often affecting all cell lines—was the main feature in all 31 available core specimens. A disproportionately pronounced effect on myeloid and megakaryocytic cells was observed in 19 cats. Myelofibrosis or bone marrow necrosis was not a feature. Conclusion and Clinical Importance Mycotoxin induced pancytopenia should be considered as differential diagnosis in otherwise healthy cats presenting with bi‐ or pancytopenia and bone marrow hypo‐ to aplasia.

recalled. Concentrations of the hemato-and myelotoxic mycotoxins T-2, HT-2 and diacetoxyscirpenol (DAS) greater than the European Commission guidance for dry cat foods were detected in the recalled diets.
Objectives: To describe clinical and clinicopathological findings in cats diagnosed with suspected diet induced pancytopenia.
Animals: Fifty cats presenting with pancytopenia after exposure to a recalled diet.
Methods: Multicenter retrospective case series study. Cats with known exposure to 1 of the recalled diets were included if presented with bi-or pancytopenia and underwent bone marrow examination.
Results: Case fatality rate was 78%. Bone marrow aspirates and biopsy examination results were available in 23 cats; 19 cats had a bone marrow aspirate, and 8 cats had a biopsy core, available for examination. Bone marrow hypo to aplasia-often affecting all cell lines-was the main feature in all 31 available core specimens. A disproportionately pronounced effect on myeloid and megakaryocytic cells was observed in 19 cats. Myelofibrosis or bone marrow necrosis was not a feature.
Conclusion and Clinical Importance: Mycotoxin induced pancytopenia should be considered as differential diagnosis in otherwise healthy cats presenting with bi-or pancytopenia and bone marrow hypo-to aplasia. Trichothecene mycotoxins (including T-2, HT-2, and diacetoxyscirpenol) belong to the largest group of Fusarium mycotoxins, and at present more than 170 different trichothecenes have been isolated. 3,4 The hemato-and myelotoxicity of trichothecenes, in particular of T-2 and HT-2 toxin but also of DAS, is well recognized in the literature, mainly attributed to their inhibitory effect on protein, RNA and DNA synthesis, induction of ribotoxic stress and apoptosis, resulting in impaired hematopoiesis in the bone marrow. [3][4][5] T-2 and HT-2 are generally associated with grain products and are most frequently and at their highest concentration found in oats and oat products. 4 Ingestion of grains contaminated with T-2 and HT-2 toxin are the underlying cause for human alimentary toxic aleukia, a disease characterized by severe bone marrow hypocellularity and high mortality rate. [6][7][8] Experimental studies in cats demonstrate the potential toxic, dose-dependent fatal effect of T-2 toxin on cats. Clinical signs before death include lethargy, anorexia, hemorrhagic diarrhea and weight loss, and postmortem examination reveals multifocal hemorrhages in several organs with markedly decreased cellularity of the bone marrow and resembling the disease in humans. [9][10][11] Despite these studies, detailed toxicological data allowing the establishment of a specific lowest observed adverse effect level (LOAEL) or no observed adverse effect level (NOAEL) for cats are lacking. 5,12 The aim of this study was to describe the clinical and clinicopathological findings of cats with suspected dietary derived mycotoxin induced pancytopenia presented during the outbreak in 2021.

| MATERIAL AND METHODS
This is a retrospective, multicenter, case series study; ethical approval was granted by the Royal Veterinary College, University of London ethics and welfare committee (URN SR2021-0148).

| Study cohort
Cats were considered to be eligible for inclusion if they presented with a neutropenia (neutrophil concentration <2.5 Â 10 9 /L) or thrombocytopenia (<150 Â 10 9 /L), or both with or without anemia (hematocrit or packed cell volume <27%) between February 1st and December 11th 2021. Additionally, cats had to have had exposure to 1 of the recalled diets as well as undergone microscopic bone marrow examination of diagnostic quality.
Data collected included date of presentation, signalment, duration and type of clinical signs, indoor/outdoor status, laboratory and body cavity imaging findings, bone marrow examination results, feline immunodeficiency, feline leukemia and feline panleukopenia virus testing results, other diagnostics and dietary information for each cat affected.

| Statistical analysis
Analyses were conducted using SPSS version 28.0 (IBM Corp). If continuous variables were non-normally distributed, they were summarized using median and range. Normally distributed, continuous variables were presented as mean and SD.

| Cat characteristics, clinical signs, and initial diagnostics
There were 26 (52%) cats that were housed exclusively indoors, and 23 (46%) that had indoor and outdoor access. Indoor/outdoor status was unknown for 1 cat. There were 19 (38%) cats from a single cat household, and 31 (62%) were from a multicat household. Of the cats from a multicat household, other cats in the household were diagnosed with bi-or pancytopenia in 19 (61%) cases. Thirteen cats (26%) were fed diet A, 23 (46%) cats diet B, and 14 cats (28%) diet C. 1 Twenty-five cats (50%) were recorded as being fed solely on a dry diet, whereas 22 cats (44%) were fed a combination of wet and dry.
The median duration of clinical signs before presentation was Other measurements are presented in Table 1.

| DISCUSSION
This study describes clinicopathological features, in particular bone marrow examination findings, of cats presenting with pancytopenia that were exposed to a diet containing the trichothecene mycotoxins  15 Estrogen toxicosis was ruled out in only 1 cat but was not suspected clinically due to a lack of compatible history in any of the cats. 16 Hemato-and myelotoxic effects of T-2 and HT-2 have been described in various species. 3 T-2 toxin affects human hematopoietic progenitor cells rather than mature, circulating blood cells and the toxic effects are attributed to inhibition of hematopoiesis. 17 Inhibition of hematopoiesis is mediated by apoptosis of CD34+ stem cells. 18 Red blood cell progenitors appear less sensitive to the toxic effect of T-2/HT-2 compared to platelet and white blood cell progenitors. 12,19 A disproportionate effect on cells of myeloid and megakaryocytic lineage, often with a left shift in the erythroid lineage, was noted in our cohort, a finding that is not typical for aplastic pancytopenia; this might be suggestive of feline erythroid precursors also being less sen- In several cats the myeloid lineage appeared to have selective disruption, specifically of the neutrophil and monocyte lineage. If myeloid precursors were present, the predominating cell type was of eosinophilic lineage. It is not possible to rule out whether those eosinophils were prominent due to a specific cytokine driven mechanism, or present in normal numbers and just not affected similarly as the other cell lineages. The presence of eosinophils on bone marrow examination is described in pigs being administered DAS but not on bone marrow specimens from dogs in the same study. 21 Pigs and dogs, both developed severe cellular necrosis in the hematopoietic cords when given DAS, and severity of necrosis appeared to be dose-related. 21 Necrosis was reported in 1 of the 50 cats in our study.
T-2 toxin is able to induce hemolysis in some animal species (eg, rabbits, dogs, horses, and humans) with variable degree of effect, whereas ruminants seem to be resistant to this effect due to degradation of T-2 toxin and related toxins by rumen microbiota. 12 Evidence of hemolysis based on mismatching hemoglobin concentration and PCV or hyperbilirubinemia was not a feature in the presented group of cats. Eight cats (16%) had mild to moderate increases in serum bilirubin concentration, though this was considered most likely due to functional cholestasis associated with severe inflammatory disease/ sepsis. 24 Diacetoxyscirpenol has a similar myelo-and hematotoxic effect compared to T-2/HT-2 and in pigs it is more toxic than T-2. 21,25 Like T-2/HT-2 it is extensively absorbed from the gastrointestinal tract and rapidly excreted via urine and feces within 12 to 24 hours. 26 The extensive metabolism of DAS is mediated by four pathways-similar to T-2/HT-2-by means of hydrolyzation, hydroxylation, deepoxidation and conjugation, with the latter pathway lacking in felids.
Cats are considered to be a sensitive species in regard to the toxic effects of T-2/HT-2 toxin which is likely associated with their inability to excrete T-2 toxin and its metabolites via glucuronide conjugation. 5 Therefore, it has to be assumed that a similar sensitivity exists toward the DAS toxin.
Case fatality rate in this study was higher compared to the case fatality rate reported for the overall affected cohort. 1 This is thought to be due to 2 main factors. One, the reported case fatality rate for the overall affected cohort is likely an underestimation due to the nature of reporting and capture of data ("1 moment only") in that study and also due to the fact that it is very likely that not all affected cats were registered on our data base. 1 In contrast, the case fatality rate reported here is likely an overestimation, as availability of bone marrow examinations results was an inclusion criterion. Bone marrow aspiration and biopsy as part of the diagnostic work-up were mainly performed on cats presenting before the food was recalled, whereas the majority of bone marrow examinations performed after the food was recalled were performed only in association with euthanasia.
After a strong link to diet had been established, cats presenting with pancytopenia and known exposure to 1 of the recalled diets were only minimally investigated and only underwent supportive treatment at the Queen Mother Hospital for Animals, Royal Veterinary College, for example, 11 of 13 cats underwent bone marrow investigations as part of their diagnostic work-up, whereas only 9 of 26 cats presenting later only underwent bone marrow aspiration/core biopsy at time of euthanasia, with 65% of the cats presenting later surviving to discharge.
A limitation of this descriptive report is that obtaining a bone marrow sample of good diagnostic quality is more challenging in cats compared to some other species such as dogs and diagnosing hypocellular bone marrows is limited when analyzing solely aspirate samples. As reports were generated by numerous pathologists, a degree of heterogeneity was present in the way individuals classified findings as hypo or aplastic and not all reports mentioned certain key features.
Due to the lack of NOAEL and LOAEL for trichothecene toxins in cats, it remains speculative whether the pancytopenia in presented cats was caused by the mycotoxins T-2/HT-2 and DAS. However, similarities in clinical presentation and clinicopathological findings, in particular bone marrow examination findings compared to other species and reported studies in cats, [9][10][11] as well as exclusion of other previously documented causes, make a causative link likely and veterinarians should consider dietary derived mycotoxin induced marrow toxicosis as a possible underlying cause in cats presenting with pancytopenia. It is interesting to note that all affected brands were marketed as grain-free, which could potentially lessen the likelihood of mycotoxin contamination on initial consideration. Occurrence of trichothecenes mycotoxins is also described in vegetable, including starchy roots vegetable like potatoes. 2,[27][28] The diets associated with the pancytopenia outbreak in cats in the United Kingdom all contained potatoes flakes which, in this scenario, are thought to have been the likely source of mycotoxin.

ACKNOWLEDGMENT
No funding was received for this study. We thank Professor J. Fink-Gremmels, University of Utrecht, for her advice regarding our description of T-2/HT-2 and DAS mycotoxins.