Lymphoma in Miniature Dachshunds: A retrospective multicenter study of 108 cases (2006‐2018) in Japan

Abstract Background Miniature Dachshunds (MD) are predisposed to lymphoma with disease onset of young age and long‐term survival. Objectives To compare clinical features and survival time of lymphoma in MD and non‐MD. Animals One hundred and eight MDs with lymphoma and 149 non‐MD breed dogs with lymphoma were included in the study. Methods This was a retrospective multicenter observational study. Lymphoma was classified based on signalment, histopathology/cytology, and anatomical site of the disease. For each type of lymphoma, median survival time was analyzed by Kaplan‐Meier estimates and life table analysis. Prognostic factors for large‐cell gastrointestinal lymphoma (LGIL) were analyzed using Cox regression. Results Gastrointestinal lymphomas were more common in MDs (53/108) compared to non‐MDs (41/149). The multicentric lymphoma was most common in non‐MD breed dogs (74/149) compared to MDs (33/108). The median age that dog developed lymphoma in MD and non‐MD were both 10 years old; however, lymphomas were more frequently observed in younger dogs (<4 years) in MDs (20/108) compared to non‐MDs (9/149; P = .002). Seventy percent were diagnosed with B‐cell with median age of diagnosis was 3 (1‐14) years. Mott cell differentiation was observed in 6 dogs. Age <4 years and B‐cell phenotype were significant factors for longer survival time in MD with LGIL. Conclusions and Clinical Importance Lymphomas in MDs involved gastrointestinal lesions at higher frequency compared to other dog breeds examined. B‐cell lymphoma was more common in early‐onset LGIL in MD and cases that involved Mott cell differentiation were observed. Awareness of this specific presentation of lymphoma in dogs will possibly affect the treatment decision process for the owners of MD with LGIL.


| INTRODUCTION
Multicentric lymphomas (ML) account for approximately 80% of cases of lymphoma in dogs; however, gastrointestinal lymphomas (GIL) account for 5% to 7%. [1][2][3][4] The median survival time (MST) for ML and GIL is 10 to 12 months and 0.5 to 2.5 months, respectively. [5][6][7][8][9] The immunophenotype of tumor cells is 1 of the reasons for the short survival time of dogs with GIL. 17,18 Among ML, B-cell lymphoma account for 70% to 80% and have a MST of 11 to 13 months, while T-cell lymphoma is chemoresistant and had a MST of 5 to 5.3 months. 5,[10][11][12] T-cell lymphoma account for 63% to 91% of GIL in dogs [6][7][8]13,14 ; the higher proportion of T-cell lymphoma might be 1 factor associated with shorter MST. Moreover, there have been a small number of cases of B-cell GIL with long-term survival. 15,16 Furthermore, colorectal lymphomas in dogs often originate in B-cells and survive long term. 17,18 A case report found that a miniature dachshunds (MD) dog had a good prognosis and different characteristics than usual and occur at a younger age (mean age = 4 years). 19 However, their study did not assess the immunological characteristics nor mentioned Mott cells. It is clinically relevant to demonstrate the difference, if any, of response to the treatment and survival time between lymphoma in MD and other breed dogs to better inform clients in MD diagnosed with lymphoma.
This study therefore aimed to compare the clinical characteristics of lymphoma in MD with those of other dog breeds.

| Study design
This was a retrospective multicenter observational study performed at 11 animal hospitals in Japan.

| Review of medical records
The age, sex, weight, anatomical site of the disease, malignancy by histopathology or cytology, immunological classification, complete remission rate, first remission, and survival time were assessed based on medical records. Lymphoma was classified as per cell morphology and size using a clinical or diagnostic pathologist. Small cell lymphoma was histopathologically diagnosed using surgical excision biopsy or endoscopic biopsy specimens. Immunological classification was performed by immunohistochemistry using CD3 or CD79 antigens or by polymerase chain reaction (PCR) for antigen receptor rearrangements (PARR). Complete remission was determined by the attending veterinarian based on physical examination results, chest and abdominal X-ray, abdominal ultrasound, and CBC results. Complete remission for GI lymphoma was determined by abdominal ultrasound when there was no lesion detected. Complete remission for multicentral lymphoma was determined by evaluating superficial lymph nodes through palpation, evaluating thoracic lymph nodes and lungs through thoracic radiograph, and evaluating abdominal lymph nodes, liver and spleen through abdominal ultrasound. As for cutaneous lymphoma, skin lesion was evaluated additional to the above examinations. First remission time was defined as the time from the day of complete remission until the confirmation of relapse. Regardless of the cause of death, survival time was defined as the period from date of diagnosis to death or euthanasia.

| Ethical statement
This was a retrospective observational study using hospital data at animal hospitals. Hence, informed consent from the animal owners was not applicable here. The study was approved by hospital boards at Saitama Animal Medical Center (Approval number: 20191001, approval date: November 15, 2019). All dogs were treated and evaluated by board-certified specialists at the hospital and provided with customary high standards of medical care and welfare.

| RESULTS
A total of 257 dogs were included in this study. One hundred eight dogs were MDs and 149 dogs were non-MDs. The breed representation for non-MDs diagnosed with lymphoma is described in Table 1 Table 2). The percentage of GIL was significantly higher in lymphoma in MD than in non-MD s (P = .0178). There was no significant difference between other anatomic sites for the MD and non-MD groups.
The difference in the presentation between MD and non-MD and substage were not determined in this study.

| Comparison of age of LGIL onset
No significant (P = .329) difference in median age of LGIL onset was seen in MD dogs (median 9 years, range 1-14) and non-MD dogs (median 9 years, range 2-14; P = .3239). However, the comparison of age distributions resulted in higher incidence of early-onset disease, and a bimodal distribution of age at onset in MD compared to the incidence age distribution in non-MDs (P = .01; Figure 1).  As such, LGIL in MD represented a significantly higher incidence of B-cell LGIL than the non-MDs (P = .001).

| LGIL onset site
The LGIL site in MD was observed in the small bowel (n = 28), No significant difference in onset localization/site was seen between MD and non-MD dogs with LGIL (P = .1).

| Treatment for LGIL
Chemotherapy was administered to 34 of 42 MD cases, of which surgery was performed before chemotherapy in 6 cases. The chemother-

| LGIL survival time
At the end of the study period, 13 of 34 LGIL cases that underwent chemotherapy were still surviving. Of 21 MDs that died, the cause of death was believed to be lymphoma in 17 MDs, cause other than lymphoma in 2 cases, and cause of death was unknown in 2 cases.
The survival time was significantly longer (P = .0001) for MD with LGIL with the MST1503 days (survival probability of 50% with 95% CI: 30-57) compared to non-MDs with the MST 56.7 days (survival probability of 50% with 95%CI: 32-65; Figure 2). Note: The number of B-cell lymphoma was significantly higher in MD than in non-MDs (P = .001).  Figure 4).

| Comparison of survival time of LGIL in MD according to immunophenotype
The survival of LGIL in MD was assessed as per immunophenotype.

| Analysis of the prognosticators of LGIL in MD
The univariate Cox regression of prognosticators of LGIL in MD resulted in age <4 years (P < .0001) and B-cell phenotype (P = .025) as significant factors (Table 3).
Lomustine and prednisolone induction therapy were administered in 7 cases, chorambucil and prednisolone in 2 cases, and COP therapy was administered in 1 case. Thirteen of the 14 non-MDs with CL were administered chemotherapy, out of which 9 were administered lomustine and prednisolone, 3 were administered chorambucil and prednisolone, and 1 was treated by CHOP.
The complete remission rate of dogs with CL that underwent chemotherapy was 20% (2/10) for MDs and 30% (4/13) for non-MDs (P = .066). All the MDs and 9 out of 14 non-MDs died in this study.

| DISCUSSION
This study demonstrated that the percentage of gastrointestinal lym- LGIL in dogs in general were characterized by a low response rate (56%) and short survival time of 0.5 to 2.5 months. 5 The limitations of this study were inherent to its retrospective design and include small sample size, variable staging and treatment regimens, as well as lack of standardized follow-up. While the population of MD was collected from multiple centers, the non-MD dogs were exclusively selected from the referral veterinary hospital, which could have resulted in sampling bias. In addition, any consequences of surgery was not examined in this study that were performed before chemotherapy.
As per the results of analyses performed in this study, lymphomas in MDs involved gastrointestinal lesions compared to other dog breeds.
The onset of LGIL in MD was bimodally distributed because onset had occurred at a young age in many dogs. Finally, B-cell lymphoma was more common in early-onset LGIL in MD and cases that involved Mott cell differentiation were observed. The veterinarian awareness of this specific presentation of lymphoma in dogs will possibly affect the treatment decision process for the owners of MD with LGIL. Aggressive chemotherapy for LGIL in MD might be effective for long-term survival.