Introduction
Clinical abnormalities secondary to hypercalcemia are common presenting complaints to small animal practices. Cancer is the leading cause of hypercalcemia in dogs, diagnosed in about 58% to 66% of dogs with hypercalcemia.1,2 The most common types of cancers causing hypercalcemia of malignancy (HM) include lymphoma and apocrine gland adenocarcinoma of the anal sac as well as thyroid carcinoma, mammary gland carcinoma, multiple myeloma, melanoma, primary pulmonary carcinomas, thymoma, and chronic lymphocytic leukemia.2–4 Nonneoplastic causes include renal disease, primary hyperparathyroidism, granulomatous disease, osteolytic disease, hypoadrenocorticism, vitamin D toxicosis, and spurious analytic results. Clinical signs of hypercalcemia are often nonspecific and depend on the severity of hypercalcemia. Multiple organ systems are affected by hypercalcemia, but the most severe clinical consequences occur in the neuromuscular, renal, cardiovascular, and gastrointestinal systems.5
The classic diagnostic approach for investigating hypercalcemia entails a thorough physical examination including screening for lymphadenopathy and a rectal examination to evaluate for an anal sac mass, laboratory work including a CBC with a blood smear examination, serum biochemistry panel, urinalysis, and parathyroid hormone (PTH) and PTH-related protein (PTH-rp) testing as well as abdominal and thoracic imaging, cervical ultrasonography, and survey bone radiographs. While there are many pathophysiologic mechanisms for HM, the most common is paraneoplastic production of PTH-rp. These dogs with HM typically have low serum PTH levels and increased PTH-rp. This pattern contrasts with dogs with primary hyperparathyroidism, which may have normal or elevated PTH levels in the face of hypercalcemia and undetectable PTH-rp.6
HM is commonly caused by soluble mediators released by tumor cells into circulation that then act on bone and kidneys through endocrine and paracrine pathways. While aberrant production of PTH-rp is the most common, various other cytokines can also contribute to the pathogenesis of humoral HM, including IL-1, IL-6, tumor necrosis factor, calcitriol, TGF-β, and receptor activator of nuclear factor kappa β ligands. Other causes of HM include ectopic production of PTH, increased serum 1,25-dihydroxyvitamin D, and extensive lytic bone metastases.6–8
Hypercalcemia is present in 57% of nodal CD4+ T-cell lymphoma cases,9 29% of nodal CD4– CD8– T-cell lymphoma cases,10 and 67% of primary mediastinal lymphoma cases, which were almost exclusively T cell in origin in dogs.11 Lymphoma as a cause of hypercalcemia is frequently diagnosed through cytology. Flow cytometry or histology can be useful to subtype lymphoma and potentially offer prognostic information.12 Flow cytometry does this by using surface markers to gain important diagnostic and prognostic information without the need for invasive procedures such as surgical biopsies.13–16 The PCR for antigen receptor rearrangements (PARR) assay is useful to assess clonality of a population of lymphocytes when there is suspicion for lymphoma that cannot be definitively diagnosed through alternative means.
In some cases, a thorough physical examination and extensive imaging may fail to reveal the cause of hypercalcemia in HM cases. We anecdotally observed 2 cases of hypercalcemia where neoplasia was not evident until the bone marrow was examined. These cases suggested a unique manifestation or entity of T-cell lymphoid neoplasia, very similar to a report of a single case of primary bone marrow T-cell lymphoid neoplasia with hypercalcemia.17 These cases prompted us to look retrospectively for additional cases of paraneoplastic hypercalcemia associated with primary bone marrow T-cell lymphoid neoplasia in dogs. We found these cases to be challenging to diagnose given the uncommon variant of lymphoid cancer present, with diagnoses only eventually established by bone marrow analysis. Lymphoid molecular diagnostics were useful to establish a definitive diagnosis and characterize the neoplasm. A subset of cases lived substantially longer than anticipated for a case presenting with T-cell neoplasia in the bone marrow, cytopenia, and hypercalcemia. The focus of this case series is to explore the clinical presentation and outcome of these cases and highlight the use of immunophenotyping in diagnosis.
Methods
Animals
This retrospective study reviewed the clinical records of dogs with hypercalcemia diagnosed with T-cell lymphoid neoplasm present within the bone marrow (by flow cytometry, with or without cytologic confirmation) without evidence of lymphoid neoplasia identified elsewhere. Two cases meeting criteria were evaluated at the Iowa State University (ISU) Veterinary Lloyd Veterinary Medical Center in 2017. The Colorado State University Clinical Hematopathology (CSU-CH) Laboratory database was queried for bone marrow samples from dogs submitted for flow cytometry between 2014 and 2021 meeting inclusion criteria. Samples were submitted from teaching hospitals (4) and tertiary hospitals (7). Case inclusion criteria included (a) ionized hypercalcemia; (b) no obvious diagnosis of cancer based on physical examination, imaging, and fine-needle aspirate of lymph nodes or organs other than bone marrow; (c) no lymphocytosis in the peripheral blood; (d) and CD34– T-cell lymphoid neoplasia diagnosed in the bone marrow by flow cytometry (with or without cytologic confirmation).
Data retrieved from medical records included patient signalment (age, sex, breed, neuter status), clinical history, presenting complaint, clinical examination findings, diagnostic results, therapeutic protocol, and outcome. Diagnostic test results including CBC, serum biochemistry profile, bone marrow aspirate cytologic evaluation with flow cytometry and PARR, abdominal and thoracic imaging, and case-relevant needle aspirate cytology of other organ systems were reviewed for all cases with available data. A blood film review was performed by a pathologist in 3 cases (Nos. 4, 6, and 9) or a medical technologist at a university or reference diagnostic lab and subsequently was not routed for a pathologist review in 5 cases (Nos. 1, 2, 5, 8, and 10). A blood film review was not reported in 3 cases (Nos. 3, 7, and 11).
Flow cytometry
Bone marrow aspirate samples were either (a) delivered immediately to the laboratory if the patient was evaluated onsite or (b) shipped to the laboratory overnight on ice and kept refrigerated until analysis. Samples were received by the laboratory and analyzed within 72 hours of being obtained from the dog. Bone marrow aspirates were collected into an EDTA solution. Routine immunophenotyping by flow cytometry was performed on a bone marrow aspirate at the CSU-CH laboratory for 9 cases and at ISU for 2 cases as previously described.16 At the CSU-CH laboratory, samples submitted prior to March 7, 2017, were stained with antibody panel 1; samples submitted between March 7, 2017, and November 13, 2019, were stained with antibody panel 2; and samples submitted after November 13, 2019, were stained with antibody panel 3 (Supplementary Table S1). All samples were analyzed on a 3-laser Coulter Gallios instrument (Beckman Coulter Inc), and data were analyzed using Kaluza Analysis Software, version 2.1 (Beckman Coulter Inc). At ISU, cases were stained using antibody panel 4 and analyzed on a FACSCanto (BD Biosciences) cytometer with data analyzed using FlowJo (BD Biosciences). Expression of class II major histocompatibility complex (MHC) on T cells was determined by median fluorescence intensity (MFI). T-cell populations were categorized as CD5– if cells had complete loss of CD5 expression, which was determined by MFI and interpretation of the flow cytometry plots. The median linear forward light scatter (FS), which is a partial reflection of cell size, was measured on neoplastic T cells. Flow cytometry data were compared to CSU-CH laboratory data from blood of 29 healthy control dogs and bone marrow of 10 separate healthy control dogs without suspicion for hematopoietic neoplasia. For each case analyzed at the CSU-CH laboratory, the ratio of the median FS of the neoplastic T-cell population to the median FS of normal CD4+ T cells from bone marrow control samples (n = 9) was calculated.
Clonality testing
The PARR assay was performed on a bone marrow aspirate as previously described.18 Two cases had PARR requested at the time of diagnosis. For remaining cases, PARR was performed retrospectively on the bone marrow aspirate submitted for flow cytometry for cases with sample material still available (n = 6).
Statistical analysis
Signalment, clinical data, and flow cytometry data were summarized. CBC and biochemistry data abnormalities were identified if values were outside the reference interval of the laboratory performing the test. Continuous data were assessed for normality by a Shapiro-Wilk test and expressed as median and range or mean and SD. Treatments with corticosteroids or chemotherapeutics were summarized. Overall survival (OS) was calculated from the date hypercalcemia was identified to the date of death or last contact. Cases that were alive at the time of data collection or lost to follow-up were censored. The Kaplan-Meier method was used to calculate median OS time using Prism, version 9.4.0 (GraphPad Software).
Results
Study population
A total of 216 unique bone marrow samples were submitted to the CSU-CH laboratory for flow cytometry between January 1, 2014, and February 20, 2021. Nine of those cases met the inclusion criteria for this study. Two bone marrow samples were also submitted to the ISU Clinical Pathology laboratory between September 9, 2017, and November 20, 2017, that met the case inclusion criteria for this study.
Case characteristics
The median age and body weight at presentation were 5.7 years (range, 4.0 to 8.6 years) and 27.85 kg (range, 17.4 to 45.2 kg), respectively (Table 1; Supplementary Table S2). Males represented 73% of cases, and 37% were Golden Retrievers. Body weight data were not available for dog 4.
Summary signalment data, laboratory data, physical examination findings, treatments, and survival data for 11 dogs with hypercalcemia and T-cell lymphoid neoplasia in the bone marrow.
| Variablea | No. (%)b of cases affected or median (range) |
|---|---|
| Age (y; n = 11) | 5.7 (4.0–8.6) |
| Body weight (kg; n = 10) | 27.85 (17.4–45.2) |
| Sex (n = 11) | |
| Male intact | 1 (9%) |
| Male castrated | 7 (64%) |
| Female spayed | 3 (27%) |
| Breed (n = 11) | |
| Golden Retriever | 4 (36%) |
| Boxer | 1 (9%) |
| Collie | 1 (9%) |
| German Shorthaired Pointer | 1 (9%) |
| Mix | 4 (36%) |
| Anemia (n = 11) | 3 (27%) |
| Thrombocytopenia (n = 11) | 10 (91%) |
| Neutropenia (n = 9) | 8 (89%) |
| Liver abnormalities (n = 9) | 3 (33%) |
| Splenic abnormalities (n = 9) | 4 (44%) |
| Visceral lymphadenopathy (n = 9) | 3 (33%) |
| Percentage of neoplastic T cells by flow cytometry (n = 11)c | 45 (9–89) |
| Treatments (n = 8) | |
| Prednisone | 1 (13%) |
| Single agent chemotherapy | 1 (13%) |
| Multiagent chemotherapy | 6 (75%) |
| Overall survival (d; n = 9) | 262 (25–792) |
| Alive/censored | 4 (44%) |
| Died | 5 (56%) |
aNumber of cases with available data.
bPercentage of cases affected among those cases with available data.
cPercentage of neoplastic T cells among all nucleated cells in the bone marrow sample by flow cytometry.
Clinical signs
The most common presenting complaint (72%) was polyuria and polydipsia, followed by hyporexia/anorexia noted in 7 of 11 dogs (64%). Other gastrointestinal signs were uncommonly reported, with 2 of 11 dogs having vomiting and 1 of 11 dogs having diarrhea. The range of time between onset of clinical signs and a visit to a veterinarian was 0 to 61 days (median, 5 days).
Initial diagnostic findings
Every dog had a physical examination, CBC, serum biochemistry panel, ionized calcium, and bone marrow aspirate for flow cytometry performed. Seven out of 11 dogs had documented rectal examinations performed, which were all normal. Three of 11 cases had a mild anemia (median Hct, 34%; range, 33% to 38%), which was considered nonregenerative in the 2 cases with an available reticulocyte count. The third case did not have a reticulocyte count reported, but the corresponding bone marrow cytology demonstrated severe marrow hypoplasia with the erythroid series described as absent, and polychromasia was not detected in the background. Lymphocyte counts were within the reference interval except in 3 cases where lymphopenia was noted. Of the 9 dogs for which neutrophil count was reported or commented upon on laboratory submission forms, 8 of 9 had neutropenia (median, 1.65 X 103/μL; range, 0.91 X 103 to 2.74 X 103/μL). Ten of 11 dogs had reported thrombocytopenia (median, 90 X 103/μL; range, 18 X 103 to 180 X 103/μL), with 8 of 11 having blood smears confirming thrombocytopenia; however, the dog reported to have a normal platelet count had no specific count number available for review. In 2 cases (Nos. 7 and 11), CBC abnormalities were obtained from the CSU-CH submission form but the CBC was not available for review. On pathologist review of case 9 there were low numbers of intermediate-sized lymphocytes, but there was not a confirmation of a neoplastic population cytologically, and additional diagnostics (flow cytometry/PARR) were not pursued on the blood to investigate for blood involvement. Of biochemistry data available, 7 of 8 dogs had azotemia (creatinine median, 2.05 mg/dL; range, 1 to 4.2 mg/dL). The median reported ionized calcium was 1.91 mmol/L (patient range, 1.44 to 2.58 mmol/L), and the median total calcium was 14.5 mg/dL (patient range, 11.7 to 17.7 mg/dL). These values were above the laboratory reference interval in all cases. In the 6 cases where PTH was measured, all had a value below or at the low end of the reference interval. Of 4 cases that had PTH-rp performed, all were undetectable. Ten of 11 dogs had an abdominal ultrasonogram that did not demonstrate any obvious masses. Three cases had subtle abnormalities/changes in echogenicity or size of the liver, and 4 cases had cytology on a liver aspirate, which was interpreted as nodular regeneration (1), mild to moderate lymphocytic infiltrate (1), mild hepatocellular rarefaction (1), and normal (1). Four had nonspecific abnormalities such as splenomegaly or splenic nodules on ultrasonographic evaluation of the spleen, and 4 cases had cytology on a splenic aspirate that revealed no abnormalities (2), reactive lymphoid hyperplasia (1), and atypical lymphoid proliferation (1). Flow cytometry and PARR were not performed on any of the liver or splenic aspirates in these cases. Three cases had mild abdominal lymphadenopathy, of which 1 was aspirated and interpreted as most consistent with mild reactive lymphoid hyperplasia. Nine dogs had thoracic radiographs; of these, 8 dogs had no evidence of thoracic lymphadenopathy or mediastinal mass, and 1 dog had possible sternal lymphadenopathy, which was not aspirated.
Bone marrow cytologic and histologic findings
Of the 9 dogs where bone marrow aspirate cytology was performed, full cytologic descriptions and interpretations were available for review. In 5 cases a diagnosis of lymphoid malignancy was made cytologically, for 3 cases bone marrow lymphocytosis with suspicion for lymphoid malignancy was reported, and for 1 case a low-density atypical lymphoid cell population concerning for lymphoid malignancy was documented. Marrow cellularity was variable between cases, with 5 of 9 being hypercellular, 2 of 9 normocellular, 1 of 9 hypocellular, and 1 of 9 having no particle units present for cellularity assessment. Iron density was reported in 4 cases, with all having normal iron amounts. Suspected hypoplasia of the megakaryocytic, erythroid, and myeloid lineages was documented in 2 dogs and dual erythroid and myeloid series hypoplasia in 3 cases. One dog had megakaryocytic hyperplasia. The 3 cases where the myeloid-to-erythroid ratio was reported had ratios of 1.2:1, 1:1, and 2:1. Bone marrow lymphocytosis was the most prominent pathologic finding reported in 8 cases, with lymphoid cells ranging from 24% to nearly 100% of nucleated cells. The remaining case had low-frequency atypical lymphoid cells. The morphology of the lymphocyte populations was mostly reported as monomorphic, in which the lymphocytes were intermediate sized (n = 4), small-intermediate sized (2), and intermediate-large sized (3). Lymphocytes had high nuclear-to-cytoplasmic ratio with round, but sometimes clefted or ameboid-shaped, nuclei. Nuclear chromatin patterns varied from finely stippled, granular, and open, with visible nucleoli reported in 5 cases. Lymphoid cell cytoplasm was reported as moderately basophilic and mildly expanded in some cases, with no cytoplasmic granules, vacuoles, or perinuclear clearing zones reported in any cases. Very-low-density histiocytic infiltrates with concurrent low-level erythrophagocytosis were reported in 2 cases. For the 5 cases in which a diagnosis of lymphoid malignancy was made cytologically, nucleoli were a reported feature of the malignant lymphocyte population; in 3 of these cases, the lymphoid population was also intermediate-large sized. In 1 of these later cases, acute lymphoblastic leukemia (ALL) was highly suspected. Bone marrow cytologic features for 2 cases are presented (Figure 1).
Bone marrow aspirate cytologic findings from 2 cases (case 1 [A and B] and case 6 [C and D]). For case 1, bone marrow was predominated by a population of small-to-intermediate sized lymphocytes (arrowhead [A and B]) that were relatively morphologically normal. A less frequent subset of large blastic lymphocytes was also present and exhibited moderate atypia, including prominent nucleoli and multiple, sometimes bizarrely shaped nucleoli (arrow [B]). Case 6 exhibited a hypercellular marrow, with intermediate-sized lymphoid cells being the majority population (arrowhead [C and D]). Atypia within the population was variable, with a subset containing round to clefted nuclei with fine chromatin, expanded cytoplasm, and perinuclear clearing zones (D).
Citation: Journal of the American Veterinary Medical Association 262, 3; 10.2460/javma.23.07.0389
Bone marrow histopathology was available for 2 cases at initial diagnosis. One case had a highly cellular core biopsy sample demonstrating myelophthisis secondary to round cell tumor. The round cells were described as large with clear cytoplasm, round nuclei, and occasional small nucleoli, and the primary differential was lymphoid neoplasia. The second case demonstrated infiltrating to nearly completely effacing cellular mass of sheets of round cells. The round cells were described as small to medium-sized with faintly eosinophilic cytoplasm with cell nuclei round and central with dense chromatin. Histologic findings were consistent with T-cell lymphoid neoplasia.
Flow cytometry
Flow cytometry on a bone marrow aspirate was performed 1 to 99 days after hypercalcemia was detected. Flow cytograms from a representative case are presented and compared to a normal bone marrow from a healthy dog (Figure 2). All cases evaluated by the CSU-CH laboratory had a population of neoplastic T cells that expressed low levels of class II MHC. Across all CSU-CH laboratory cases (Nos. 3 to 11), the median class II MHC MFI of neoplastic T cells was 0.5 (IQR, 0.4 to 1.7; range, 0.2 to 5.1), compared to a median of 20.9 (IQR, 19.7 to 23.0; range, 16.8 to 33.2) in normal CD4+ T cells from control bone marrow samples (n = 10) and a median of 17.9 (IQR, 14.7 to 20.1; range, 11.7 to 27.2) in normal CD4+ T cells from control blood samples (29). In the 11 cases reported in this study, the neoplastic T cells accounted for 9% to 89% of all nucleated cells in the bone marrow sample by flow cytometry. In control bone marrow from healthy dogs (n = 10), CD3+ T cells account for 3% to 20% of total nucleated cells (median, 7%; IQR, 6% to 13%). In all cases, the neoplastic T cells expressed the T-cell antigen CD3 and the pan-leukocyte antigen CD45, and T cells did not express the stem cell antigen CD34. In 9 cases the neoplastic T cells expressed CD4, in 1 case the neoplastic T cells did not express the subset antigens CD4 or CD8, and in 1 case (No. 10) a subset of neoplastic T cells expressed CD4 and a subset did not express CD4 or CD8. Six of 11 cases did not express the T-cell antigen CD5. Among CSU-CH laboratory cases, approximate cell size was assessed by comparison of the median FS of the neoplastic T-cell population by flow cytometry to the median FS of normal CD4+ T cells from bone marrow control samples (n = 10). Neoplastic T cells were 1.0 to 1.4 times the size of control CD4+ T cells.
Flow cytometry and PCR for antigen receptor rearrangement (PARR): flow cytometric immunophenotyping of neoplastic CD4+ T cells in the bone marrow of a dog (A) and normal CD4+ T cells in the bone marrow of a healthy control dog (B). Size plots on the left with linear forward light scatter on the horizontal axis and with log side light scatter on the vertical axis. Fluorescence dot plots in the middle show that CD4+ T cells (green) express the antigens CD4, CD3, and CD5. Neoplastic CD4+ T cells in the case (A) are small to intermediate in size. Neoplastic CD4+ T cells express low levels of class II MHC (A) compared to normal CD4+ T cells from healthy bone marrow (B; right plots). PARR results of bone marrow are shown from a dog with CD4+ T-cell lymphoid neoplasia (C) and a control dog with no suspicion for neoplasia by cytology (D). T-cell receptor γ rearrangements (blue and black) are on the left, and immunoglobulin rearrangements (green) are on the right. Amplicon sizes are displayed on the horizontal axis, and relative fluorescence (a measure of amplicon abundance) is on the vertical axis. C—In the case with hypercalcemia and CD4+ T-cell lymphoid neoplasia detected in the bone marrow by flow cytometry (A), the results demonstrate clonal T-cell receptor rearrangements and polyclonal immunoglobulin rearrangements, supporting the diagnosis of T-cell lymphoid neoplasia. D—In the control dog with no suspicion for lymphoid neoplasia, results demonstrate polyclonal T-cell receptor and immunoglobulin rearrangements.
Citation: Journal of the American Veterinary Medical Association 262, 3; 10.2460/javma.23.07.0389
Clonality testing
Eight of 11 cases had sample material available for clonality testing by the PARR assay. Seven of 8 cases had a clonal T-cell receptor rearrangement detected (Figure 2), and 1 case had a suspicious T-cell receptor rearrangement peak that did not reach objective criteria for clonality. Immunoglobulin rearrangements were polyclonal for all 8 cases.
Treatment and outcome
Information on treatment was available for 8 of 11 cases (Table 1; Supplementary Table S2). One case was treated with prednisone, 3 cases (Nos. 5 to 7) were treated with a CHOP-based protocol, and the remaining 4 cases (Nos. 1 to 3 and 10) were treated with variable chemotherapeutic agents with or without prednisone. In 4 cases (Nos. 2, 5, 6, and 7), the clinician documented relapse 201 to 269 days (median, 237 days) after initial diagnosis. For cases in which parameters of relapse were documented, the definition included development of pancytopenia (1), peripheral lymphadenopathy (1), neurologic signs (1), and recurrence of thrombocytopenia and hypercalcemia (1). Rescue therapies were pursued in 2 of 4 and included a repeated CHOP-based protocol (case 7) and L-asparaginase (case 5), although case 5 was euthanized shortly afterward. For those that did not receive treatment at relapse, one was euthanized due to the progression of disease (case 2), and one died before treatment was started (case 6). For the cases in which parameters of treatment response were documented, all 6 cases had clinical improvement as response to treatment; 2 of 6 used resolution of hypercalcemia, and 1 case used improvement in neutropenia. Follow-up information was available for 9 of 11 cases. Of those 9 cases with outcome data available, 5 patients died (70 to 293 days after initial diagnosis of hypercalcemia), and 4 were lost to follow-up (25 to 792 days after initial diagnosis). The median OS was 260 days.
Necropsy findings
Two cases underwent necropsy evaluation. One case (No. 2) was euthanized shortly after relapse was documented, 234 days after diagnosis, and the necropsy demonstrated disseminated lymphoid neoplasia in the bone marrow, liver, spleen, and lymph nodes despite the aspirates of the liver and spleen demonstrating no evidence of neoplasia initially. The other case (No. 1) demonstrated no evidence of lymphoid neoplasia or any other cancer types on necropsy, was undergoing treatment with doxorubicin, and was euthanized because of concern for progression of lymphoma due to development of peripheral lymphadenopathy.
Discussion
This retrospective case series describes hypercalcemic dogs with T-cell lymphoid neoplasia diagnosed by bone marrow evaluation that may represent a unique primary bone marrow T-cell lymphoid neoplasm causing paraneoplastic hypercalcemia. These cases frequently had neutropenia and thrombocytopenia, but lacked peripheral lymphocytosis or overt evidence of lymphoma in solid tissues and required investigation into the bone marrow to obtain a diagnosis. The treatments and survival times were variable across cases, but a subset of these cases had an unexpectedly good outcome given the presence of hypercalcemia and cytopenias, consistent with a prior case report17 of a similar entity. This series highlights the utility of examining the bone marrow in cases with unexplained hypercalcemia and cytopenias that lack overt evidence of lymphoma/leukemia elsewhere. Additionally, this series highlights the utility of flow cytometry to definitively diagnose neoplasia in a subset of cases and characterize the neoplasm as a CD4+ or CD4– CD8– T-cell neoplasm, rather than a CD34+ acute leukemia, which may be suspected clinically given the cytopenias.
The cases reported here represent an unusual presentation that does not clearly fall into any established lymphoid neoplasia classification as defined by the WHO.19 The flow cytometry features of these neoplasms, having predominantly CD4+ CD45+ T cells that express low levels of class II MHC, are similar to those of nodal CD4+ peripheral T-cell lymphoma. However, these suspected primary bone marrow cases appeared to lack nodal involvement and often survived longer than expected for peripheral T-cell lymphoma.9 In addition, although this neoplasm affected the bone marrow and there are cytopenias present, these T cells do not express CD34, and a subset of cases had prolonged survival (up to 792 days), which is inconsistent with the aggressive clinical course of ALL (median survival time, 9 days; range, 1 to 120 days) and suggests this is a distinct entity from ALL.20 Although thrombocytopenia and neutropenia were seen in a high proportion of our cases similar to ALL, only 3 of 11 cases had anemia, all of which were mild, and none of these dogs had a peripheral lymphocytosis. This finding contrasts with ALL where 1 study demonstrated that 90.6% of cases had anemia and 70.3% had a lymphocytosis20 and a second study reported that 91% of all acute leukemias (acute myeloid leukemia and ALL) were anemic.21 A recent study22 reported a CD4– CD8– T-cell lymphoid neoplasm with similar flow cytometric parameters (small cells that expressed low class II MHC levels) to cases described in this series. However, that disease had a very different clinical presentation in that it affected young dogs, had a predilection for English Bulldogs, had cytologically evident circulating neoplastic lymphocytes, and commonly had hepatic and gastrointestinal signs, in addition to the fact that hypercalcemia was uncommon. A recent case report17 demonstrated a T-cell lymphoid neoplasm restricted to the bone marrow with similar flow cytometry characteristics (small to intermediate-sized CD4+ CD3+ CD5– CD34– CD45+ T cells), neutropenia, and hypercalcemia, which survived 16.5 months after diagnosis. In 9 of 11 cases in our study where data were available, the T cells appeared to express lower class II MHC and dogs had a more variable outcome, but many aspects of the clinical presentation were similar. All the cases in this series had small to intermediate-sized neoplastic T cells by flow cytometry. The cell size description was variable by cytology and histology, and we suspect discrepancies in cell size estimations are attributed to differences in methodology.23–26
Flow cytometry on the bone marrow was useful in a subset of these cases to diagnose neoplasia where cytology was not definitive. When the lymphocytes are small and not substantially expanded, it can be challenging to distinguish a reactive from neoplastic process by cytology, but the aberrant phenotype detected by flow cytometry allows for a definitive diagnosis even if the proportion of neoplastic cells is low. Flow cytometry was also useful to characterize the neoplasm as CD4+ or CD4– CD5– T cell in origin, rather than a CD34+ ALL, which was suspected clinically in some cases given the infiltration in the marrow and cytopenias and suspected cytologically in 1 case (No. 5), which had a survival time of 293 days. The cell of origin of these tumors is unclear, but it is possible these are tumors of mature naive T cells given the expression of CD4 and low class II MHC. These results suggested that flow cytometry of the bone marrow should be considered in cases in which there is an unexplained hypercalcemia after extensive laboratory testing and diagnostic imaging, particularly if bone marrow cytology indicates lymphocytosis in the marrow or an atypical lymphoid population. Furthermore, since most cases had both thrombocytopenia (91% [10/11 cases]) and neutropenia (80% [8/10 cases]), such dual cytopenias can be used as indications to include a bone marrow aspirate cytology and flow cytometry as part of a hypercalcemia workup. The cause of the cytopenias in these cases was not always clear. In the 2 cases with bone marrow biopsies, myelophthisis was evident, but in the bone marrow cytology samples, erythroid and myeloid series ranged from hypocellular to hypercellular. Other possible mechanisms for the peripheral cytopenias include an immune-mediated destructive process, marrow necrosis or fibrosis, or hormonal and/or metabolic-associated paraneoplastic effects.27,28
Though T-cell lymphoid neoplasms frequently cause hypercalcemia through PTH-rp, the 4 cases in this series that had PTH-rp testing performed did not have detectable PTH-rp, suggesting other mechanisms of HM occurred in these cases. These 4 cases also highlight the importance of not ruling out lymphoid neoplasia based on a normal PTH-rp. Other possible mechanisms of HM include other inflammatory mediators, increased serum vitamin-D family levels, or osteolysis. These other factors may also work synergistically or additively with PTH-rp.6 Although we cannot preclude osteolysis, there were no indications on physical examination or radiographic evidence of osteolytic lesions. (1,25-(OH)2D) levels were not measured in any cases, so this cannot be precluded as a cause or contributing factor.
The survival times ranged significantly between our reported cases (25 to 792 days), despite the similarities in clinical presentation and flow cytometry features. The variation in survival could be reflective of the specific treatments pursued, different courses of disease, or presentation at different stages of disease, or the variation in survival could represent different clinical entities. Additionally, nearly half of the cases were alive at the time of data collection or were lost to follow-up, so the full range of survival times could not be determined in this retrospective series. One of the cases (No. 10) had a higher class II MHC expression than others (though still considered low as compared to normal T cells) and had the longest survival time prior to loss of follow-up (792 days). This may argue against these cases representing a single type of T-cell lymphoid neoplasia, but individual response variation occurs in dogs with high-grade multicentric lymphoma,29 which is partly due to the biology of malignant cells undergoing major genetic and epigenetic changes during their growth and proliferation.30 Hypercalcemia and the presence of cytopenias are often considered evidence for more aggressive forms of malignancy, so it is important to note that a subset of cases with T-cell lymphoid neoplasia, hypercalcemia, and cytopenias have a favorable prognosis.
Limitations of this study include lack of standardized and complete diagnostic workup, including a PTH/PTH-rp/Calcitriol panel, CBC with pathologist review, abdominal ultrasonogram, thoracic radiographs, fine-needle aspirate cytology of liver and spleen, and hepatic and splenic clonality testing or flow cytometry. Three cases did not have a reported blood film review by a medical technologist or clinical pathologist, which precludes the ability to completely exclude the presence of circulating neoplastic lymphocytes. Four cases had liver and/or splenic aspirates, where neoplasia was not identified; however, it is possible the viscera could have had a low-density infiltrate of neoplastic lymphocytes that were below the level of cytologic detection and diagnosis. Not all cases had thoracic and abdominal imaging, and not all cases had liver and spleen sampled, so we cannot definitively conclude that the lymphoid neoplasm was confined to the bone marrow, but from the physical examination and diagnostics performed, there was no overt evidence of neoplasia elsewhere. The decision to aspirate the liver and spleen, especially if they appear normal on ultrasonographic examination, to investigate a hypercalcemic dog is not standard across clinicians and may be contraindicated in a patient with significant thrombocytopenia. Not every case had serum PTH concentrations measured, which could have helped exclude the rare possibility of dual primary hyperparathyroidism and marrow-centric T-cell lymphoid neoplasia. Rectal examinations were performed in only 7 of 11 cases, which precludes the possibility of concurrent apocrine gland anal sac carcinoma adenocarcinoma. Only 2 of 11 cases had necropsies performed, of which 1 demonstrated progressive disease from the sole original location in the bone marrow.
Additional limitations of this study include its retrospective nature, lack of long-term follow-up in all cases, variable treatments, and small number of cases. Further work is needed to better characterize this potentially unique presentation of T-cell lymphoid neoplasia and determine whether it is a distinct entity in dogs. Particularly it would be useful to assess cases with standardized diagnostic testing, to rule out disease more definitively in other sites, and collect more complete treatment and outcome data. This series highlights the utility of investigating the bone marrow in cases of hypercalcemia, when the cause cannot be deduced from lab work, imaging, and extramedullary tissue sampling. Additionally, flow cytometry is useful to offer a definitive diagnosis, characterization of the neoplasm, and distinction from a CD34+ ALL. In summary, this case series highlights a potentially unique presentation of primary bone marrow T-cell lymphoid neoplasia that warrants further investigation and consideration during a hypercalcemic diagnostic investigation.
Supplementary Materials
Supplementary materials are posted online at the journal website: avmajournals.avma.org
Acknowledgments
The authors would like to thank the veterinarians and clients that provided clinical information for this study as well as the laboratory personnel who performed the flow cytometry and PCR for antigen receptor rearrangement assays.
Disclosures
Drs. Rout and Avery are employed by the Colorado State University Clinical Hematopathology Laboratory, which offers diagnostic tests, including flow cytometry and clonality testing (PCR for antigen receptor rearrangement), on a fee-for-service basis. The authors have no other conflicts of interest to disclose.
No AI-assisted technologies were used in the generation of this manuscript.
Funding
The authors have nothing to disclose.
References
- 1.↑
Uehlinger P, Glaus T, Hauser B, Reusch C. Differential diagnosis of hypercalcemia—a retrospective study of 46 dogs. Article in German. Schweiz Arch Tierheilkd. 1998;140(5):188-197.
- 2.↑
Messinger JS, Windham WR, Ward CR. Ionized hypercalcemia in dogs: a retrospective study of 109 cases (1998-2003). J Vet Intern Med. 2009;23(3):514-519. doi:10.1111/j.1939-1676.2009.0288.x
- 3.
Kohart NA, Elshafae SM, Breitbach JT, Rosol TJ. Animal models of cancer-associated hypercalcemia. Vet Sci. 2017;4(2):21. doi:10.3390/vetsci4020021
- 4.↑
Robat CS, Cesario L, Gaeta R, Miller M, Schrempp D, Chun R. Clinical features, treatment options, and outcome in dogs with thymoma: 116 cases (1999-2010). J Am Vet Med Assoc. 2013;243(10):1448-1454. doi:10.2460/javma.243.10.1448
- 5.↑
Elliott J, Dobson J, Dunn JK, Herrtage ME, Jackson KF. Hypercalcaemia in the dog: a study of 40 cases. J Small Anim Pract. 1991;32(11):564-571. doi:10.1111/j.1748-5827.1991.tb00887.x
- 6.↑
Rosol TJ, Nagode LA, Couto CG, et al. Parathyroid hormone (PTH)-related protein, PTH, and 1,25-dihydroxyvitamin D in dogs with cancer-associated hypercalcemia. Endocrinology. 1992;131(3):1157-1164. doi:10.1210/endo.131.3.1505457
- 7.
Guise TA. Molecular mechanisms of osteolytic bone metastases. Cancer. 2000;88(suppl 12):2892-2898. doi:10.1002/1097-0142(20000615)88:12+<2892:AID-CNCR2>3.0.CO;2-Y
- 8.↑
Stewart AF. Hypercalcemia associated with cancer. Clinical Practice. N Engl J Med. 2005;352(4):373-379. doi:10.1056/NEJMcp042806
- 9.↑
Avery PR, Burton J, Bromberek JL, et al. Flow cytometric characterization and clinical outcome of CD4+ T-cell lymphoma in dogs: 67 cases. J Vet Intern Med. 2014;28(2):538-546. doi:10.1111/jvim.12304
- 10.↑
Harris LJ, Rout ED, Labadie JD, et al. Clinical features of canine nodal T-cell lymphomas classified as CD8+ or CD4-CD8- by flow cytometry. Vet Comp Oncol. 2020;18(3):416-427. doi:10.1111/vco.12568
- 11.↑
Moore EL, Vernau W, Rebhun RB, Skorupski KA, Burton JH. Patient characteristics, prognostic factors and outcome of dogs with high-grade primary mediastinal lymphoma. Vet Comp Oncol. 2018;16(1):E45-E51. doi:10.1111/vco.12331
- 12.↑
Valli VE, Kass PH, San Myint M, Scott F. Canine lymphomas: association of classification type, disease stage, tumor subtype, mitotic rate, and treatment with survival. Vet Pathol. 2013;50(5):738-748. doi:10.1177/0300985813478210
- 13.↑
Comazzi S, Gelain ME. Use of flow cytometric immunophenotyping to refine the cytological diagnosis of canine lymphoma. Vet J. 2011;188(2):149-155. doi:10.1016/j.tvjl.2010.03.011
- 14.
Comazzi S, Riondato F. Flow cytometry in the diagnosis of canine T-cell lymphoma. Front Vet Sci. 2021;8:600963. doi:10.3389/fvets.2021.600963
- 15.
Seelig DM, Avery P, Webb T, et al. Canine T-zone lymphoma: unique immunophenotypic features, outcome, and population characteristics. J Vet Intern Med. 2014;28(3):878-886. doi:10.1111/jvim.12343
- 16.↑
Seelig DM, Avery AC, Ehrhart EJ, Linden MA. The comparative diagnostic features of canine and human lymphoma. Vet Sci. 2016;3(2):11. doi:10.3390/vetsci3020011
- 17.↑
Pinard CJ, Ludwig L, Egan R, et al. Primary bone marrow T-cell lymphoma in a Golden Retriever. Vet Clin Pathol. 2021;50(1):142-150. doi:10.1111/vcp.13009
- 18.↑
Colopy LJ, Shiu KB, Snyder LA, Avery AC, Rout ED, Moore AR. Immunoglobulin G4-related disease in a dog. J Vet Intern Med. 2019;33(6):2732-2738. doi:10.1111/jvim.15624
- 19.↑
Swerdlow SH, Campo E, Pileri SA, et al. The 2016 revision of the World Health Organization classification of lymphoid neoplasms. Blood. 2016;127(20):2375-2390. doi:10.1182/blood-2016-01-643569
- 20.↑
Novacco M, Comazzi S, Marconato L, et al. Prognostic factors in canine acute leukaemias: a retrospective study. Vet Comp Oncol. 2016;14(4):409-416. doi:10.1111/vco.12136
- 21.↑
Harris RA, Rout ED, Yoshimoto JA, Avery PR, Avery AC. Using digital RNA counting to establish flow cytometry diagnostic criteria for subtypes of CD34+ canine acute leukaemia. Vet Comp Oncol. 2022;20(3):710-719. doi:10.1111/vco.12825
- 22.↑
Frankhouse KA, Rout ED, Hughes KL, et al. An aggressive CD4- CD8- T-cell neoplasm in young English Bulldogs. Vet Comp Oncol. 2022;20(2):416-426. doi:10.1111/vco.12787
- 23.↑
Gong JZ, Williams DC Jr, Liu K, Jones C. Fine-needle aspiration in non-Hodgkin lymphoma: evaluation of cell size by cytomorphology and flow cytometry. Am J Clin Pathol. 2002;117(6):880-888. doi:10.1309/16UL-W4PX-HRL7-7V88
- 24.
Gelain ME, Mazzilli M, Riondato F, Marconato L, Comazzi S. Aberrant phenotypes and quantitative antigen expression in different subtypes of canine lymphoma by flow cytometry. Vet Immunol Immunopathol. 2008;121(3-4):179-188. doi:10.1016/j.vetimm.2007.09.018
- 25.
Bernardi S, Martini V, Perfetto S, Cozzi M, Comazzi S. Flow cytometric analysis of mediastinal masses in cats: a retrospective study. Front Vet Sci. 2020;7:444. doi:10.3389/fvets.2020.00444
- 26.↑
Rout ED, Fernandez M, Yoshimoto JA, Hughes KL, Avery AC, Burton JH. Clinical outcome and Ki67 evaluation in dogs with nodal small cell B-cell lymphoma diagnosed by flow cytometry. J Vet Intern Med. 2022;36(5):1770-1781. doi:10.1111/jvim.16515
- 27.↑
Weiss DJ, Evanson OA, Sykes J. A retrospective study of canine pancytopenia. Vet Clin Pathol. 1999;28(3):83-88. doi:10.1111/j.1939-165X.1999.tb01053.x
- 28.↑
Grindem CB, Breitschwerdt EB, Corbett WT, Page RL, Jans HE. Thrombocytopenia associated with neoplasia in dogs. J Vet Intern Med. 1994;8(6):400-405. doi:10.1111/j.1939-1676.1994.tb03258.x
- 29.↑
Marconato L, Stefanello D, Valenti P, et al. Predictors of long-term survival in dogs with high-grade multicentric lymphoma. J Am Vet Med Assoc. 2011;238(4):480-485. doi:10.2460/javma.238.4.480
- 30.↑
Savage P, Stebbing J, Bower M, Crook T. Why does cytotoxic chemotherapy cure only some cancers? Nat Clin Pract Oncol. 2009;6(1):43-52. doi:10.1038/ncponc1260
