Introduction

Bone Marrow Aspirate (BMA) is an essential tool for studying, staging, and monitoring various conditions involving the bone marrow in veterinary medicine (^{Javinsky, 2012}; ^{Stacy and Harvey, 2017}). This technique offers a reliable means of obtaining accurate diagnoses and is relatively safe, with minimal risks and few complications (^{Woods et al., 2021}). The effective use of BMA in routine veterinary practice requires proper sample collection, accurate smear interpretation, and comprehensive reporting that provides clear diagnostic information (^{Messick, 2023}).

The BMA report is a crucial component of the post-analytical stage and significantly influences the veterinarian's understanding and decision-making process (^{Hawkins, 2012}; ^{Sciacovelli et al., 2016}). A thorough report must include essential elements such as sample traceability, patient identification, correlation with complete blood count, and clinical history or prior diagnostic tests. Additionally, it should compare findings with reference values and, when necessary, suggest additional tests such as immunophenotyping, cytochemistry, cytogenetics, and molecular biology. These components are critical for differential diagnosis and a deeper understanding of bone marrow findings (^{Stacy and Harvey, 2017}).

Advancements in human and veterinary hematology have prompted the development of standardized guidelines aimed at optimizing processes from test requests to result interpretation and reporting (^{Riley et al., 2021}; ^{Ritt, 2022}). In veterinary medicine, these guidelines frequently draw from human models, leveraging physiological similarities across species such as dogs, cats, and humans (^{Comazzi et al., 2017}; ^{Rütgen et al., 2022}).

In Colombia, research on BMA in companion animals remains scarce, leading to the underestimation of key factors related to its application in dogs and cats, whose combined population exceeded eleven million in 2022 (^{Ministerio de Salud y Protección Social, 2022}).

The objective of this study was to describe the zoogeographic, clinical, and quality characteristics of BMA reports, as well as the frequency of diagnoses and associated factors, in dogs and cats treated in veterinary centers in Colombia between 2012 and 2023.

Materials and Methods

Results

A total of 139 veterinary centers across 19 provinces (25 municipalities) in Colombia were contacted. Of these, 32% (45/139) offered BMA collection, reading, or both. However, half of these outsourced the service. Out of the remaining 22 institutions, eight agreed to participate. Most of these were pathology laboratories that received samples from 46 veterinary centers in capital municipalities of seven provinces (Table 1).

A total of 135 BMA reports were obtained; 73.3% of these were reported between 2018 and 2023. The dataset included 116 dogs and 19 cats, with an average age of 5.2 ± 3 years (dogs 5.5 ± 3.2 years, median 6; cats 3.1 ± 3.0 years, median 2). Of these, 53.3% were male.

Among the patients with consultations (n = 83/135; 61.4%), 75/83 (90%) presented hematological alterations, 17/83 (20%) had a history of hemotropic agent infections, and 41/83 (49%) exhibited clinical signs (Table 2).

Additional findings included musculoskeletal abnormalities, respiratory signs, ascites, and petechiae. Blast cells were reported in three blood smears, and one cat tested positive for feline leukemia virus.

The aspirate samples were mostly taken from live patients (130/135; 96.3%), but the sample collection site was reported in only 36/135 (26.6%) of cases. The reported collection sites were the sternum (n=1), femur (n=4), humerus (n=27), ilium (n=2), and tibia (n=2).

A total of 39 dog breeds were identified. The most frequent breeds (≥3 individuals) included Beagle, Golden Retriever, French Poodle, Labrador, German Shepherd, Shih Tzu, Springer Spaniel, Yorkshire Terrier, Schnauzer, Pinscher, Pitbull, and mixed breeds. Among cats, nine breeds were identified, including Bombay, British Shorthair, Domestic Shorthair, European Shorthair, Balinese, Russian Blue, Exotic, and Persian. However, 8/19 (42%) were mixed breeds.

The distribution of dog breeds by size was: small (22/116; 19%), medium (69/116; 59%), and large (24/116; 21%). Additionally, adults of both species were the most common age group (Table 3). Additionally, adults of both species were the most common age group (Table 3).

Factors associated with diagnoses

Granulocytic hypoplasia was more common in females (9/44; 20.5%) than in males (2/54; 3.7%; p = 0.011). However, neoplasms were more common in males (16/54; 29.6%) than in females (4/44; 9.1%; p = 0.013), and erythroid hyperplasia was only present in males (6/54; 11.1%; p = 0.031; Table 7). Hypoplasia occurred more frequently in dogs (30/83; 36.1%) than in cats (1/15; 6.7%; p = 0.032). However, erythroid hyperplasia was more common in cats (3/15; 20.0%) than in dogs (3/83; 3.6%; p = 0.044). Granulocytic hyperplasia was more frequent in small-breed individuals (4/13; 30.8%; p = 0.013). There was no breed predominance in cats for any diagnosis, nor was there in the disaggregated dog breeds. Four of the patients who presented with medullary aplasia had pancytopenia (4/9; 44.4%; p = 0.002). Other individuals with pancytopenia were diagnosed with hypoplasia, myelodysplasia, and Ehrlichia spp. infection.

Table 7 Factors associated with bone marrow diagnoses in dogs and cats from Colombia (2012-2023). 

Discussion

The total number of BMA reports obtained during the study period from eight institutions differs significantly with findings reported by ^{Weiss (2006}a, ^2006b) and ^{Turinelli et al. (2015}, 2018). Weiss (2006a) compiled 203 samples from cats and 717 from dogs in a single American institution over eight years. Similarly, Turinelli et al. (2015) gathered 295 dog samples over two years and 152 cat samples over three years in a European laboratory. These comparisons suggest that BMA is less commonly performed in Colombia than in other countries. Additionally, prior studies have reported that bone marrow evaluations are more frequent in dogs than in cats, a trend consistent with our study, where the dog-to-cat ratio was 8:1. This discrepancy in Colombia may be attributed to the higher population of dogs compared to cats (^{Ministerio de Salud y Protección Social, 2022}). Furthermore, given that hematological alterations in cats are frequently associated with highly prevalent viral infections, BMA may not be routinely performed as a differential diagnostic tool (^{Molina, 2020}; ^{Ortega et al., 2020}; ^{Santisteban et al., 2021}).

Persistent anemia, either alone or accompanied by other hematological abnormalities, was the most common indication for BMA, followed by thrombocytopenia. This aligns with the findings of ^{Turinelli et al. (2015}), where 37.6% of patients underwent BMA for persistent anemia. This hematological condition is frequently encountered in veterinary practice and is often associated with solid tumors, hematopoietic malignancies, inflammatory diseases, immune-mediated disorders, and renal dysfunction (^{Chervier et al., 2012}; ^{Grzelak and Fry, 2022}). Therefore, BMA is crucial for the differential diagnosis of non-regenerative, persistent anemia of unknown origin (^{Raskin and Messick, 2012}).

Several BMA reports lacked essential information such as patient sex, age, clinical history, marrow particle description, cellularity assessment, morphological evaluation, and differential count. These omissions hinder the proper interpretation of BMA findings. Experts in human hematopathology have demonstrated the advantages of implementing structured reports, which gather both essential and analytical elements, avoiding unstructured narrative formats (^{Sever et al., 2016}). This approach facilitates the analysis of the results against established diagnostic criteria, especially considering that the final report directly impacts the decision-making of the treating physician (^{Mylonakis and Hatzis, 2014}; ^{Stacy and Harvey, 2017}).

As previously described, sample quality is a well-known limiting factor in the analytical process of BMA (^{Trejo-Ayala et al., 2015}). This study excluded 27.4% of samples due to quality issues. This finding aligns with ^{Turinelli et al. (2015}, ²⁰¹⁸), who reported a percentage of non-diagnostic samples in dogs and cats of 30.5% and 26.5%, respectively. However, these rates contrast with ^{Weiss (2006}a, 2006b), who reported significantly lower non-diagnostic sample rates of 2.5% (cats) and 1.6% (dogs).

Hemodilution has been described as the primary cause of BMA rejection, and deficiencies associated with this error can be mitigated through technical improvements, such as particle selection and proper smear preparation (^{Riley et al., 2021}). However, the responsibility for the sampling process lies primarily with the clinician, as demonstrated by ^{Siddon et al. (2021}).

The most frequent diagnosis in dogs was bone marrow hypoplasia (36.1%), a rate closely aligned with findings by ^{Girardi et al. (2017}) in Brazil, where 26.2% of 65 dogs with pancytopenia presented with hypoplasia. However, this rate differs from the findings of Weiss (4.3%) and ^{Turinelli et al. (7.8%) (Turinelli et al., 2015}; Weiss, 2006b). This group of abnormalities has been linked to medication use, chemical substances, radiation, infections, immune disorders, and neoplasms in both humans and animals (^{Grimes and Fry, 2015}; ^{Javinsky, 2012}; ^{Moore and Krishnan, 2023}). A more detailed investigation into the causes of these abnormalities was not possible due to the limited clinical history information available for the patients.

Moreover, hyperplasia was the second most frequent diagnosis, with a prevalence of 15.7% in dogs, a rate similar to those previously reported in the United States (20.2%) and Europe (25.1%) (^{Turinelli et al., 2015}; ^{Weiss, 2006}b). Non-neoplastic hyperplasia must be correlated with clinical findings and diagnostic aids to determine the underlying causes, as it may be a temporary response to hemolytic anemias, chronic inflammatory processes, or other factors (^{Haines et al., 2022}; ^{Orazi et al., 2006}).

Despite the small sample size, neoplasms were the most frequent diagnoses in cats. No single cell line predominated, resulting in similar proportions of lymphoproliferative, myeloproliferative, and myelodysplastic disorders. For instance, ^{Gilroy et al. (2011}) documented a case of chronic eosinophilic leukemia in a male cat with persistent eosinophilia, mirroring findings in our study. Similarly, ^{Campbell et al. (2013}) described 18 cases of chronic lymphocytic leukemia in cats over a ten-year period, while ^{Patel et al. (2005}) identified 16 cases of multiple myeloma in cats over an eight-year period. Both studies applied diagnostic criteria to both peripheral blood samples and BMAs, revealing a higher number of cases than those found in our study, despite covering a longer period. This discrepancy suggests a potential underdiagnosis of marrow conditions in dogs and cats.

Additionally, vector-borne diseases have been associated with various hematological abnormalities, often manifesting as reductions in one or more cell lines (^{Bonilla-Aldana et al., 2022}; ^{Tommasi et al., 2014}). This corresponds with our study's results, where eight patients with myelodysplasia, aplastic anemia, and hypoplasia had evidence of Ehrlichia spp., Babesia spp., and Anaplasma spp. in BMAs, which had not been previously detected in peripheral blood samples.

A higher frequency of males was observed in the neoplasm group, which differs from ^{Turinelli et al. (2015}, 2018), who reported similar sex distributions in malignant cases. Other studies in dogs have suggested a greater predisposition in males to various neoplasms; however, specific data regarding sex differences in hematopoietic neoplasms are scarce (^{Pinello et al., 2022}). A study of 3,400 dogs showed that males tend to be diagnosed with cancer at an earlier age than females (^{Rafalko et al., 2023}); however, in the present study, no association between age and diagnosis was found, despite most patients being adults at the time of aspiration.

Nine cases were identified in which females presented with granulocytic hypoplasia. It has been established that the bone marrow of dogs is highly sensitive to suppression caused by endogenous or exogenous estrogens (^{Sontas et al., 2009}; ^{Weiss, 2022}). However, clinical history data were insufficient to identify whether treatments or underlying conditions may have contributed to this disorder.

No publications discussing sex-related predisposition in dogs or cats were found for erythroid hyperplasia; in this study, all cases occurred in males. This condition has been associated with peripheral causes of cell destruction and myelodysplastic neoplasms (^{Haines et al., 2022}; ^{Newman and Stokol, 2022}; ^{Weiss and Aird, 2001}).

In addition, alterations in the megakaryocytic line were reported, but there were no descriptions of the use of BMAs in patient follow-up, lymphoma staging, or the application of molecular biology techniques, cytogenetics, microbiological cultures, and immunophenotyping in any of the reports as complementary studies for the final diagnosis. However, these techniques are commonly used in veterinary medicine to define lineage and clonality and establish prognoses (^{Evans, 2023}; ^{Rout et al., 2019}). It should be noted that the availability of these tests is limited in Colombia.

To our knowledge, this is the first publication on the characterization of BMA reports in dogs and cats in Colombia. Therefore, it is likely the first report on the frequency of diagnoses in BMAs for both species. The study of BMA as a diagnostic tool in dogs and cats treated at participating veterinary centers in the country remains uncommon. The most frequent diagnosis in dogs was hypoplasia, while neoplasms were more common in cats, considering the sample size.

This study highlights the need to adhere to BMA reporting guidelines to improve diagnostic accuracy, shorten turnaround times, and avoid reprocessing. Likewise, sampling and slide preparation techniques must be optimized to reduce the number of particle-free and hemodiluted samples.

The use of BMA proved effective in diagnosing and ruling out central hematologic alterations in dogs and cats with multiple clinical and laboratory findings. Moreover, this tool enables differential diagnoses in cases where the BMA result is normal, thereby supporting a peripheral origin.