Individualized chemotherapy strategy in early-stage canine diffuse large B-cell lymphoma: a case report

Article information

Korean J Vet Res. 2026;66.e6

Publication date (electronic) : 2026 May 28

doi : https://doi.org/10.14405/kjvr.20260003

Seongwon Heo, Jihyeon Go, Seha Ahn, Joon Woo Lee, Dong-In Jung, DoHyeon Yu, Kyu-Woan Cho, Hyeona Bae^,

Department of Veterinary Internal Medicine, College of Veterinary Medicine, Gyeongsang National University, Jinju 52828, Korea

^*Corresponding author: Hyeona Bae Department of Veterinary Internal Medicine, College of Veterinary Medicine, Gyeongsang National University, 501 Jinju-daero, Jinju 52828, Korea Tel: +82-55-772-2378 E-mail: vetbha@gnu.ac.kr

Received 2026 January 13; Revised 2026 April 5; Accepted 2026 April 20.

Diffuse large B-cell lymphoma (DLBCL) is the most prevalent canine lymphoma, typically presenting as multicentric disease with complete nodal effacement (World Health Organization [WHO] stages III–V) [1]. Conversely, early-stage presentations (WHO stages I–II) account for less than 10% of cases and pose unique diagnostic challenges [2]. A primary challenge in early-stage disease is partial effacement, in which neoplastic infiltration is insufficient to completely disrupt the underlying lymph node architecture [2]. This can lead to inconclusive results via false-negative or inconclusive findings on standard diagnostic tools, such as fine needle aspiration (FNA) and flow cytometry (FC), thereby necessitating molecular confirmation through polymerase chain reaction for antigen receptor rearrangement (PARR) [3].

This report details a case of early stages of DLBCL, where the findings were most compatible with the diagnosis of DLBCL was achieved by integrating multiple diagnostic modalities. Furthermore, it illustrates that standard-of-care regimens may require patient-centered adjustments, highlighting the utility of a diagnostic algorithm and tailored treatment intensity according to individual patient characteristics.

An eleven-year-old mixed-breed dog presented with submandibular lymphadenomegaly. Following a right-sided single lymphadenectomy at a local hospital, histopathology revealed widespread large lymphocyte infiltration; however, partial effacement and preservation of residual normal architecture precluded a definitive histological diagnosis, yielding findings only suspicious for early-stage lymphoma (Fig. 1A and B).

Fig. 1.

Histopathological findings of the right submandibular lymph node. (A) Normal nodal architecture is extensively effaced and expanded by dense sheets of neoplastic cells (hematoxylin and eosin [H&E] stain, scale bar = 50 μm). (B) A population of densely packed, monomorphic neoplastic cells is present. These cells are intermediate to large (7.5–9 μm), with well-defined borders, round to oval nuclei, coarsely stippled chromatin, and prominent nucleoli (H&E stain, scale bar = 50 μm). (C) Proliferative follicle-like structures show diffuse, strong nuclear immunopositivity, confirming the B-cell lineage of the neoplastic cells (asterisks) (Pax5 stain, scale bar = 50 μm). (D) Moderate numbers of immunopositive T-cells are observed in the zones surrounding the proliferative follicular structures (arrow) (CD3 stain, scale bar = 50 μm).

Subsequent physical examination identified a left submandibular lymph node measuring 3.0 × 2.2 cm. While diagnostic imaging and baseline labs were unremarkable, further characterization was sought via immunohistochemistry (IHC), FC, and PARR. A sample from the remaining left submandibular lymph node was obtained and subjected to FC and immunophenotyping.

IHC confirmed B-cell lineage (PAX5-positive) infiltrating the follicles (Fig. 1C) with moderate numbers of surrounding CD3-positive T-cells (Fig. 1D). While FC identified (Fig. 2) a CD21^⁺/CD79a^⁺ B-cell population, the absence of clear phenotypic aberrancies or a dominant outlier clone rendered the FC findings non-diagnostic for malignancy. However, PARR detected a definitive monoclonal rearrangement of the immunoglobulin heavy chain (IgH2). Collectively, these findings most compatible with a diagnosis of stage II DLBCL. However, because bone marrow assessment was not performed, a limitation in complete staging remains [1,2].

Fig. 2.

Flow cytometric analysis of the left submandibular lymph node. Analysis reveals a heterogeneous population consisting of B-cells and T-cells with a ratio of approximately 6:4. SCC, side scatter; FSC, forward scatter.

Prior to treatment, an ex vivo chemosensitivity assay (ImpriMed Inc., USA) was performed to experimentally assess the predicted efficacy of individual chemotherapeutic agents, and vincristine and prednisolone showed particularly high efficacy [4]. Because the agents included in the CHOP (cyclophosphamide, hydroxydaunorubicin, oncovin, prednisone) protocol demonstrated favorable results overall, a 19-week CHOP protocol was initiated as planned [1].

In the first week, vincristine (0.7 mg/m²) and oral prednisolone were administered. Treatment response was monitored using the Veterinary Cooperative Oncology Group (VCOG) consensus criteria, with the sum of the longest diameters (sum of LD) measuring 31 mm [5]. Adverse effects were monitored according to the VCOG Common Terminology Criteria for Adverse Events (VCOG-CTCAE v2) [6].

One week after treatment initiation, physical examination and complete blood count (CBC) revealed a complete response (CR) accompanied by grade 4 neutropenia. Chemotherapy was delayed for 1 week until neutropenia improved. Subsequently, cyclophosphamide (250 mg/m²) was administered. On the subsequent visit, grade 1 neutropenia was noted, and the planned vincristine dose was postponed accordingly. Following the administration of doxorubicin (25 mg/m²) as the fourth agent in the first treatment cycle, the patient developed gastrointestinal and systemic complications, including grade 2 vomiting, grade 2 diarrhea, hematochezia, and pancreatitis, that required hospitalization.

Although CR was maintained on serial evaluations with no clinical evidence of relapse, recurrent chemotherapy-associated adverse events, most notably neutropenia and doxorubicin-associated gastrointestinal toxicity, necessitated modification of the treatment approach to allow continuation of the therapy. The therapy was modified to vincristine monotherapy, administered every 3 weeks, based on a prior chemosensitivity assay, which identified vincristine as the most effective agent. The only complication observed was neutropenia, which was clinically managed.

After modification of the treatment protocol, two additional doses of vincristine were administered; however, chemotherapy was discontinued because neutropenia persisted. Five months later, the patient remained in CR with no evidence of relapse. Neutropenia is clinically asymptomatic; hence, no additional intervention has been instituted, and the patient is being monitored with periodic follow-up examinations and CBC assessments.

Histopathology and IHC remain the gold standards for diagnosing canine lymphoma, providing the necessary framework for grading, therapeutic selection, and prognostic staging [3,7]. However, cases involving a single lymph node are uncommon, resulting in a paucity of data regarding early-stage disease [2].

In most high-grade lymphomas, neoplastic lymphocytes diffusely infiltrate and obliterate normal lymph node structures [2,3,7]. Contrastingly, early-stage disease or specific subtypes may exhibit residual lymphoid follicles that mimic reactive hyperplasia, complicating the histopathological interpretation [2,3,7]. Consequently, histopathology alone may be insufficient, necessitating adjunctive molecular testing.

While FC and PARR are complementary, their utility is influenced by the method of sample acquisition and the extent of tumor infiltration [3,7]. FC allows for the characterization of diverse immunophenotypes and detection of aberrant antigen expression; however, it relies on suspended cells obtained via blind FNA [7]. In cases of early-stage lymphoma, lack of a dominant neoplastic population can render FC results inconclusive or indistinguishable from reactive hyperplasia [7].

Contrastingly, PARR assesses molecular clonality by detecting antigen-receptor gene rearrangements [7]. It is used to confirm the presence of cancer in lesions that are morphologically or immunophenotypically ambiguous and serves as a useful tool for distinguishing neoplastic and reactive conditions [7]. In early-stage lymphoma, the preserved nodal architecture likely diluted the neoplastic population in the FC sample. Under these circumstances, PARR is particularly useful because it can detect a clonal lymphoid population even when tumor burden is low. In the present case, the preceding diagnostic tests failed to reveal clear differences between reactive hyperplasia and lymphoma, whereas PARR provided the molecular evidence needed to distinguish the two conditions. Therefore, although FC is preferred for lineage assignment when feasible, PARR serves as a definitive diagnostic modality in the specific context of ambiguous immunophenotypic presentation [7].

Although CHOP-based multi-agent chemotherapy is the established strict adherence to high-intensity regimens may not be optimal for all patients [1]. Standard chemotherapy is typically administered at the maximum tolerated dose; however, this intensity may be excessive in patients with early-stage disease or high chemosensitivity, leading to unnecessary toxicity [8]. Most key studies have focused on dogs with advanced-stage disease (stages III–V); hence, evidence-based protocols for early-stage lymphoma are lacking, necessitating a more cautious, individualized approach [9].

Chemotherapy-induced toxicities are common in canine oncology and often necessitate treatment delays [8]. Such postponements typically do not worsen prognosis; the occurrence of certain adverse effects can sometimes correlate with improved outcomes [8,10,11]. However, if adverse events persist, it may be necessary to reduce the dose, modify the protocol, or discontinue treatment altogether [8]. In such cases, dose reduction according to individual tolerance may help preserve antitumor efficacy while minimizing treatment-related toxicity [8].

When standard protocols fail owing to intolerance, treatment decisions may require individualized adjustment beyond established recommendations.

Herein, rapid CR was achieved with CHOP; however, persistent treatment-limiting toxicities prevented continuation of the original regimen. Therefore, the transition to vincristine monotherapy should not be interpreted as a guideline-based or literature-supported recommended modification of the standard protocol. Rather, it was a pragmatic and individualized empirical choice undertaken under exceptional clinical circumstances, with vincristine selected because the assay identified it as the agent with the highest predicted sensitivity and a comparatively favorable tolerability profile in this dog. Nevertheless, this individualized approach should be interpreted cautiously, as the early remission may also have reflected the underlying biology of the tumor itself.

Historically, single-agent protocols have not demonstrated superiority over multi-agent regimens [1]. Although vincristine was used in this case, no published studies or case reports evaluating its antineoplastic efficacy as a stand-alone monotherapy are available. In contrast, doxorubicin has been extensively evaluated as a single agent; although it achieves CR rates comparable to those of multi-agent regimens, it results in shorter response and survival times [9]. While multi-agent regimens remain the mainstream standard of care, single-agent therapy may be a potential alternative with further research or the development of novel therapeutic agents.

Although CHOP-based protocols are widely regarded as the most effective treatments for canine DLBCL and are associated with favorable outcomes, they may not apply uniformly to all dogs. In such exceptional cases, selected single-agent therapies may offer clinical benefits. In the present case, although complete remission was initially achieved with the CHOP protocol, subsequent treatment with vincristine monotherapy was selected based on ex vivo chemosensitivity testing, prior clinical response, and tolerability considerations. Nonetheless, the apparent success of this de-escalated approach should be regarded as hypothesis-generating only, as a single case with limited follow-up does not permit practice-informing conclusions regarding treatment de-escalation. Currently, robust evidence supporting individualized drug selection and management strategies for early-stage lymphoma remains limited. Consequently, prospective studies with larger cohorts are necessary to further investigate the clinical presentation and therapeutic plans in cases similar to those described in our current report.