Clinical Use of Lymphocyte Subset Analysis: As a Prognostic Marker for Dogs with Pyometra


 BackgroundLymphocyte subset analysis is clinically applied in human medicine. However, lymphocyte subset analysis is rarely used in small animal practice. We hypothesized that lymphocyte subsets analysis was useful in small animal practice as a biomarker for evaluating immune competence and predicting the disease prognosis. Lymphocyte subset analysis was performed prospectively for pyometra, a common disease in dogs, to assess its clinical usefulness in small animal practice.ResultsThis study included 29 dogs diagnosed with pyometra. They were classified into group 1 and group 2 on the basis of clinical course postoperatively. Sixteen dogs were classified in group 1 with no adverse events postoperatively. Thirteen dogs experienced adverse events such as increase in C-reactive protein concentration and white blood cell count, discharge from operation site, and hypoglycemia. These dogs were classified as group 2. Nine dogs were below the reference interval for the lymphocyte subset, eight of which were in group 2. Group 2 included significantly more dogs with lymphocyte subset abnormalities (p = 0.005). In the multivariable logistic regression analysis, only the result of lymphocyte subset analysis was significantly associated with adverse events (p = 0.02, 95% confidence interval = 1.68–192). Most dogs in group 2 were successfully treated.ConclusionsThese results indicate that lymphocyte subset analysis is useful as a prognostic tool for pyometra. Further studies are necessary for evaluating the clinical usefulness of lymphocyte subset analysis in pyometra and other diseases.


Plain English Summary
Althoughlymphocyte subset analysis is commonly applied in human clinical medicine, it is rarely used in small animal practice. We aimed to demonstrate that lymphocyte subset analysis is also useful for small animals. We performed lymphocyte subset analysis for 29 dogs diagnosed with pyometra, a common canine disease. Of these 29 dogs, 13 experienced adverse events postoperatively. We then compared dogs with or without adverse events to understand the factors that may affect the occurrence of adverse events. The results revealed that dogs with any lymphocyte subset depletion were signi cantly more prone to adverse events. Eight out of nine dogs with a decreased lymphocyte subset experienced adverse event. This nding supports that lymphocyte subset analysis is useful as a prognostic tool for pyometra. Further studies are necessary to evaluate the usefulness of lymphocyte subset analysis in small animal practice.
Background Page 3/14 Lymphocyte subset analysis is used to classify lymphocyte components by surface antigen. The clinical usefulness of lymphocyte subset analysis has been reported in human medicine. It is useful in the diagnosis of lymphoid tumors and immunode ciency. CD4 + T-cell count is associated with the mortality of patients with human immunode ciency virus infection and the infection-related indicators of systemic lupus erythematosus (SLE) (1,2). In patients with acute myeloid leukemia, lymphocyte subset analysis is useful in the differentiation of speci c subtypes and prediction of prognosis (3). In human patients with SLE, peripheral CD4 + T-cell count was revealed to be negatively correlated with serum C-reactive protein (CRP) concentration, and patients with CD4 + T-cell depletion are more prone to develop infections (1).
Lymphocyte subset abnormalities may correlate with decreased immune competence in humans. In recent years, lymphocyte subset analysis for dogs has become possible in a commercial laboratory in Japan. Some studies have also evaluated lymphocyte subsets in dogs. Decreased lymphocyte subsets due to aging or speci c diseases (4-7) and immunode ciency due to congenital B-cell depletion (8,9) have been reported. Another study proposed that decreased lymphocyte subsets may be useful as a prognostic indicator of mortality (10). However, lymphocyte subset analysis is not often used clinically in small animal practice.
Pyometra is a common disease in intact bitches and affects approximately 19%-25% of all intact bitches before 10 years of age (11,12). Pyometra has a good prognosis if appropriate treatment is provided.
However, some dogs experience a prolonged treatment duration (12). It has been reported that a closed cervix associated with more severe illness and that leukopenia is a risk factor for peritonitis and prolonged hospitalization (12,13). However, not all dogs with a prolonged treatment duration show a closed cervix or leucopenia. The etiology of pyometra is not fully understood. A previous study suggested that inhibition of mitogen-driven lymphocyte proliferation and suppression of immune system activity may occur in dogs with pyometra (14). Thus, we hypothesized that the prognosis of canine pyometra is affected by the suppression of immune system activity. Lymphocyte subset abnormalities are associated with reductions in immune competence in both veterinary and human medicine. Therefore, lymphocyte subset abnormalities may be associated with adverse secondary complications of pyometra. To evaluate its clinical usefulness, we performed lymphocyte subset analysis in dogs with pyometra and evaluated the association between the results and adverse secondary complications postoperatively for canine pyometra.

Patient population
Dogs diagnosed with pyometra at Yuki Animal Hospital between July 2018 and July 2021 were prospectively included. All dogs underwent physical examinations as well as CBC, plasma biochemistry pro le analysis, thoracic and abdominal radiography, and abdominal ultrasonography. If pyometra was suspected from these analyses, ovariohysterectomy and cytological examination of a uid sample from the uterus were performed. Clinical diagnosis of pyometra was made based on these clinical ndings (22). Bacterial culture of uid, blood culture, and arthrocentesis was performed when a clinical diagnosis of pyometra was made. In this study, dogs with underlying diseases were not excluded. This study was designed to evaluate the association between the presence of underlying disease and adverse events. Bacterial culture was outsourced to a commercial laboratory (Animal Medical Technology, Aichi, Japan).
Since laboratory results typically take 5-7 days, bacterial culture was also performed using the disk method at our hospital to rapidly obtain sensitivity test results.
Postoperatively, antibacterial treatment was implemented under hospitalization. CBC and CRP concentrations and, in some cases, other plasma biochemistry pro les, were monitored daily. Antibacterial drugs were started during the perioperative period. The initial antibiotic was selected from ampicillin or cephalexin (27). If the clinician deemed it necessary, enro oxacin was used in combination with these antibiotics. Adequate antibacterial drugs were selected depending on the results of the inhospital bacterial culture. If the outsourced and in-hospital sensitivity test results differed, the antibacterial drugs were changed in accordance with the outsourced result. Antibacterial drugs were continued until suture removal or the CRP concentration decreased to the reference range (<0.9 mg/dl). The dogs were discharged at the clinician's discretion when continuous decrease of CRP concentration for >2 days postoperatively and improvement of appetite and activity were con rmed. Adverse events were de ned as any abnormal clinical presentations such as diarrhea, discharge from operation site, hypoglycemia, and re-increase in CRP concentration or WBC count between surgery and suture removal. If CRP concentration and WBC count were increased from the previous day, they were judged to be reincreased. Monitoring was continued until the CRP concentrations decreased to within the reference interval or the adverse events were cured. Dogs with arthritis were not given immunosuppressive therapy and rechecked for arthrocentesis at the time of suture removal.
Dogs diagnosed with pyometra were classi ed into group 1 or group 2 depending on the clinical course postoperatively. Dogs that experienced no adverse events until suture removal were classi ed into group 1, whereas dogs that experienced any adverse event were classi ed into group 2. The history, signalment, laboratory test results, treatment duration, and lymphocyte subset analysis results were compared between the two groups. Lymphocyte subset analysis was re-performed in dogs that showed a reduction in any lymphocyte subset at the time of re-examination after treatment was completed, as long as the owner's permission was obtained.

Blood collection and quanti cation
We collected blood samples for CBC, plasma biochemistry pro le analysis, lymphocyte subset analysis, and blood bacteriologic culture analysis from all dogs via venipuncture of the cephalic, saphenous, or jugular vein. Samples were placed in tubes with or without an anticoagulant. Plasma was separated from blood within 30 min of collection. These samples were analyzed on the day of collection. For blood bacteriologic culture, 1 mL of whole blood was collected in bottles for aerobic culture (Versa TREK™ REDOX™ 1 EZ Draw™; Thermo Fisher Scienti c, Waltham, MA, USA) and anaerobic culture (Versa TREK™ REDOX™ 2 EZ Draw™; Thermo Fisher Scienti c, Waltham, MA, USA).

Assays
We measured the CBC using an automated hematology analyzer (IDEXX ProCyte Dx; IDEXX Laboratories, Westbrook, MA, USA). We obtained the plasma biochemical pro le by using a dry chemistry analyzer (Fuji lm DRI-CHEM 7000 V; Fuji lm Corporation, Tokyo, Japan). An in-hospital antibacterial susceptibility test was performed by the disk diffusion method. Purulent uid collected from the uterus was inoculated directly in Mueller Hinton agar plates and antibacterial disks (KB disk, EIKEN CHEMICAL CO., LTD., Tokyo, Japan). After incubation at 37°C for 48 h, the zone diameters were measured. The standard of susceptibility depended on the package insert.

Lymphocyte subset analysis
Lymphocyte subset analysis was outsourced to the Animal Allergy Clinical Laboratories (AACL; Kanagawa, Japan). We collected 1 mL whole blood specimens mixed with EDTA on the same day as ovariohysterectomy. These blood specimens were sent to AACL within 2 days after collection. The measurements of the same samples by an examiner were within 5% for T-cells and within 3% for the other cells. The error between the examiners ranged from 2% to 3%. Reference intervals were established from the data of 86 healthy dogs. The median age of these healthy dogs was 4 years (range, 6 months to 14 years).

Statistical analyses
The CBC and biochemical analysis results, age, body weight, days from onset to visit in our hospital, and length of hospitalization were compared using the Mann-Whitney U test. The presence or absence of anorexia, underlying diseases, lymphocyte subset analysis abnormalities, and arthritis, along with the results of bacterial culture and blood culture, were compared using Fisher's exact test. Multivariable logistic regression analysis including the results of lymphocyte subset analysis and other variables potentially associated with adverse events was performed. Variables with p-values of <0.05 were considered statistically signi cant. Statistical analyses were performed using the software Easy R (28).

Lymphocyte subset analysis
Lymphocyte subset abnormalities were found in 9 dogs. The results of lymphocyte subset analyses are presented in Table 1. Cases 1, 3, and 5 were re-examined with lymphocyte subset analysis after hospital discharge. In case 1, both the B-and helper T-cell (Th-cell) counts remained low on days 23 and 133. In cases 3 and 5, the T-cell count increased to within the reference interval on days 70 and 53, respectively. Eight out of 9 dogs classi ed as group 2.
in Table 2. The median treatment duration in group 2 was signi cantly longer than that in group 1 (p = 0.003). Group 2 included signi cantly more dogs with lymphocyte subset analysis abnormalities (p = 0.005). The frequency of underlying disease was higher in group 2 than in group 1, but this was not statistically signi cant (p = 0.09). In a multivariable logistic regression analysis including age, lymphocyte subset analysis abnormalities, and underlying disease, only the presence of lymphocyte subset analysis abnormalities was signi cantly associated with adverse events (p = 0.02, 95% con dence interval = 1.68-192). Underlying diseases included myxomatous mitral valve disease (American College of Veterinary Internal Medicine class B2) (n = 3), myxomatous mitral valve disease (American College of Veterinary Internal Medicine class B1) (n = 1), mammary gland tumor (n = 3), chronic kidney disease (International Renal Interest Society stage 3) (n = 1), bladder tumor (n = 1), hyperadrenocorticism (n = 1), and hypothyroidism (n = 1). Most group 2 dogs were successfully treated, and treatment was mainly with antibacterial therapy. Hypoglycemia was treated with glucose infusion. One dog that experienced severe nonregenerative anemia received a blood transfusion.
Comparison of clinicopathological results between the normal and abnormal groups No signi cant differences were found between group 1 and group 2 in all parameters of the complete blood count (CBC) and plasma biochemical pro le. Blood culture was positive in 1 dog in group 1 and 3 dog in group 2 (p = 0.60). Arthritis was found in 5 dogs in group 1 and 2 dog in the abnormal group (p = 0.40). All cases of arthritis resolved 10 days postoperatively without immunosuppressive therapy. Isolated bacteria did not vary signi cantly between the two groups (p = 0.89) ( Table 3).

Discussion
In the present study, signi cantly more dogs with decreased B-and/or T-cell at diagnosis of pyometra experienced complicated outcomes. CRP concentration is generally known to rapidly decrease to within the reference interval on the 10th day after surgical treatment for pyometra (15). The treatment duration of group 2 was signi cantly longer than that of group 1. These results indicate that the decreased Bor/and T-cell counts may be negative prognostic markers of pyometra. As the CRP concentrations in group 2 were decreased with changes in or increased doses of the antibiotic drugs, the re-increase in the CRP concentration was considered to be caused by persistent infection, although no infection site was found in the postoperative examination. In human medicine, increased susceptibility to infection is caused by a decline in lymphocyte counts under the condition of some diseases and aging (2,(16)(17)(18).
Several studies have been also conducted in dogs about decreases in B-and T-cell counts. The decrease of B-or T-cell counts have been reported in dogs with common variable immunode ciency, chronic kidney disease, and diabetes mellitus (5,6,8,9). However, these diseases were not found among the cases in the present study. Negative correlations between age and the number of B-and T-cells were reported in healthy beagles in a previous study (4). In the present study, although data are not shown, the median age of dogs with decreased B-or/and T-cells (12 years and 7 months) was signi cantly higher than that of the other dogs (8 years and 1 month) (p = 0.05). An age-related increase in susceptibility to infections may develop in dogs, as it does in humans. However, age was not associated with the occurrence of adverse events. This result suggests that adverse event cannot be predicted by age alone.
In cases 3 and 5, the T-cell percentage increased after treatment. In human medicine, proteinenergy malnutrition (PEM) is a well-known factor of immune-compromised condition by starvation (19). It leads to increased susceptibility to infection. However, only few studies have researched on PEM in dogs despite the fact that nutritional status is also believed to affect the immunity of the animal. In a study with mice, lymphocytes, especially CD4 + T-cells, were reduced in the mice fasted for a period as short as 48 h (20). Another study showed that healthy cats fasted for 4 days have reduced CD4 + T-cells, which was reversed by refeeding (21). These previous studies suggested that animals as well as humans who go through starvation may develop PEM. Cases 3 and 5 had anorexia for 1 week and 2 weeks, respectively. Although not supported by this study, the temporary decrease in the percentage of T-cells in these two cases could be caused by acute starvation. The decreased immune competence caused by starvation may worsen the prognosis for pyometra.
We evaluated other factors that may be involved in immune response and prognosis. Endotoxemia, sepsis, and reactive polyarthritis are associated with pyometra (22,23). Although not proven, they are believed to potentially worsen prognosis. However, these factors appear to have a low impact on the lymphocyte subset analysis and prognosis.
Some adverse events in the present study are not considered speci c to pyometra. For example, one dog that experienced severe nonregenerative anemia was diagnosed with hyperadrenocorticism, hypothyroidism, and chronic kidney disease as underlying diseases. Nonregenerative anemia may have been caused by these underlying diseases. This dog showed no abnormalities on lymphocyte subset analysis. It should be noted that dogs with underlying diseases were not excluded in the present study. In addition, dogs with pyometra tend to be old. As a result, underlying disease is often diagnosed at the same time as pyometra. Although it was expected that underlying diseases may affect prognosis, no signi cant association was found. Furthermore, infection of the surgical site and increased WBC count, CRP concentration, and liver enzyme activity may be caused by the surgical procedure or anesthesia itself. If lymphocyte subset analysis can predict these postoperative complications, it may also be useful in predicting prognosis not only for pyometra but also for other diseases that require surgical treatment.
The present study has some limitations. First, the lymphocyte subset analysis is susceptible to errors during ow cytometry analysis. The method used for ow cytometric analysis of peripheral blood lymphocytes has been described previously (24,25). In the present study, the gating technique and cell surface markers used were similar to these previous studies mentioned. Second, whether the prognosis of pyometra should be evaluated using the numbers or percentages of lymphocytes is arguable. Prognosis was di cult to determine using the number of each lymphocyte component because the number of lymphocytes varied greatly between the individual cases. We consider that the use of percentages is a useful and simple method. Third, the timing of sampling differed from case to case because of differences in the period from onset and data from follow-ups. Lastly, we did not perform histopathological examination and immunohistochemistry of the uterus. De nitive diagnosis of pyometra is made based on postoperative macroscopic and histopathological examination ndings of the uterus as well as microbiological examination of uterine contents (22). However, pyometra can be strongly suspected through the combination of clinicopathological ndings and diagnostic imaging without histopathological examination (22,23). We clinically diagnosed pyometra by comprehensively evaluating the results of all tests we performed. When immunohistochemistry of the uterus is performed, local changes in lymphocyte subsets may be present (26). Since we are a primary care animal hospital, this examination could not be performed due to a lack of equipment.

Conclusion
Lymphocyte subset analysis has been shown to be useful as a tool for predicting the prognosis of canine pyometra. This study demonstrates that lymphocyte subset analysis may be useful for predicting prognosis not only for pyometra but also for other diseases. The usefulness of lymphocyte subset analysis in small animal practice should be further evaluated in cases of other disease and pyometra.