ACEi: Angiotensin-converting enzyme inhibitor; ACVIM: American College of Veterinary Internal Medicine; Ao: Aorta; CPE: Cardiogenic pulmonary edema; CRI: Continuous rate infusion; LA: Left atrium; LVIDdN: Left ventricular end  diastolic diameter normalized for body weight; MMVD: myxomatous mitral valve disease; NT-proBNP: N-terminal pro-B-type natriuretic peptide; VHS: vertebral heart size; VLAS: vertebral left atrial size.

Pulmonary edema is a life-threatening pathological condition that can arise from either cardiogenic or non-cardiogenic causes (Mattu et al., 2005). Non-cardiogenic pulmonary edema develops independently of cardiac disease and may result from increased local hydrostatic pressure, alterations in alveolar or capillary permeability, or a combination of both mechanisms (Mattu et al., 2005). In contrast, cardiogenic pulmonary edema (CPE) occurs due to increased pulmonary venous hydrostatic pressure secondary to elevated left atrial pressure, leading to excessive interstitial fluid accumulation beyond the capacity of lymphatic drainage (Glaus et al., 2010).
       
Myxomatous mitral valve disease (MMVD) is the most common cardiac disorder in dogs and is characterized by progressive left atrial enlargement, congestive heart failure, pulmonary edema and sudden death (Borgarelli and Buchanan, 2012; Borgarelli et al., 2008; Kumar and Kumar, 2024; Yadav et al., 2023). In 2019, the American College of Veterinary Internal Medicine (ACVIM) proposed a staging system (Stages A-D) based on structural and functional cardiac abnormalities and provided stage-specific treatment guidelines. For life-threatening pulmonary edema, recommended therapies include oxygen supplementation, diuretics such as furosemide, vasodilators including angiotensin-converting enzyme inhibitors (ACEIs) and positive inotropic agents such as pimobendan (Keene et al., 2019).
       
Furosemide, the most commonly used diuretic in clinical practice, reduces circulating blood volume, thereby decreasing cardiac preload as well as pulmonary capillary and venous pressures, ultimately alleviating pulmonary edema (Atkins and Häggström, 2012). In cases of severe respiratory distress accompanied by labored breathing and frothy sputum, furosemide may be administered as a bolus; however, continuous rate infusion (CRI) is indicated when the initial response is inadequate. Excessive administration of furosemide can result in severe dehydration, necessitating careful monitoring of patient status and appropriate dose adjustment (Keene et al., 2019). In dogs presenting with severe dyspnea and suspected CPE, diuretics are sometimes administered as a therapeutic trial. Improvement of pulmonary edema can be assessed through clinical signs and thoracic radiographic findings.
       
Differentiation between cardiogenic and non-cardiogenic pulmonary edema is commonly performed using radiographic assessment of cardiac size, including vertebral heart size (VHS) and vertebral left atrial size (VLAS) (Buchanan and Bücheler, 1995; Lord et al., 2010; Malcolm et al., 2018; Saini et al., 2023). Both VHS and VLAS are useful indicators for predicting asymptomatic cardiac enlargement corresponding to ACVIM Stage B2 in dogs with MMVD (Duler et al., 2021). VHS has been extensively studied as a parameter for evaluating cardiac size across various dog breeds (Bodh et al., 2016) and numerous studies have investigated the clinical utility of VHS and VLAS in the diagnosis and management of cardiac disease (Wesselowski et al., 2022). However, in dogs with CPE, frequent diagnostic imaging may not always be feasible in critically ill dogs. In addition, correlations between radiographic cardiac indices and biomarkers such as NT-proBNP and C-reactive protein have further supported the clinical utility of VLAS in dogs with MMVD (Chae et al., 2021; Hwang and Song, 2021). To date, no studies have evaluated the relationship between changes in body weight and radiographic cardiac size before and after diuretic therapy in dogs with cardiogenic pulmonary edema. Therefore, an objective and easily measurable parameter that reflects treatment response without repeated radiographic examinations would be clinically valuable.
       
Maltese dogs represent a large proportion of the companion dog population in Korea and are a small-breed dog with known risk factors for MMVD; consequently, several breed-specific studies have been reported (Baisan et al., 2021; Baisan and Vulpe, 2022; Lee et al., 2019; Tsai et al., 2021). Recent studies have also suggested that Maltese dogs may possess a genetic predisposition to MMVD (Lee et al., 2019).
       
Therefore, this study aimed to investigate whether body weight reduction occurring during furosemide treatment is associated with radiographic improvement in Maltese dogs with cardiogenic pulmonary edema and to evaluate its potential clinical utility as an adjunctive monitoring parameter.

Animals
 
This study retrospectively reviewed medical records of patients presented to the Daejeon Time Animal Medical Center, Korea between June 2020 and June 2022. Among Maltese dogs diagnosed with pulmonary edema, those treated with continuous rate infusion (CRI) of furosemide were identified. All dogs fulfilled the diagnostic criteria for ACVIM Stage C myxomatous mitral valve disease according to the 2019 ACVIM Consensus Guidelines. A total of 118 Maltese dogs were initially diagnosed with pulmonary edema during the study period and relevant baseline information, hematologic data and imaging findings were obtained from electronic medical records (Woorien; pnV, Seoul, Korea). All medical records were used with informed consent obtained from the owners prior to clinical treatment.
 
Inclusion criteria
 
Maltese dogs presenting on an emergency basis with dyspnea as the chief complaint, diagnosed with myxomatous mitral valve disease (MMVD), confirmed to have cardiogenic pulmonary edema based on physical examination and diagnostic imaging and treated with furosemide CRI were included in the study. Diagnosis of cardiogenic pulmonary edema was based on compatible clinical signs, thoracic radiographic findings, echocardiographic evidence of MMVD with left-sided cardiac enlargement and exclusion of other causes of pulmonary infiltrates.
 
Exclusion criteria
 
Dogs diagnosed with pulmonary edema but showing evidence of pneumonia or pulmonary thromboembolism based on clinical signs, hematologic examination, or imaging findings (n = 46), as well as dogs diagnosed with non-cardiogenic pulmonary edema (NCPE; n = 24), were excluded from the analysis.
 
Sample collection
 
Blood samples were collected from the cephalic vein and placed into EDTA-3K tubes and lithium heparin tubes (Catalyst lithium heparin whole blood separator) for anticoagulation. To assess patient status, complete blood count (CBC), serum biochemical analysis (Chem 17 CLIP) and cardiac biomarker measurement (NT-proBNP) were performed.

Hematologic and biochemical analysis
 
Only blood samples without coagulation, hemolysis, or degradation were included in the analysis. Complete blood counts were performed using EDTA-3K–anticoagulated whole blood and an automated hematology analyzer (ProCyte, IDEXX Laboratories, Westbrook, ME, USA). Parameters measured included red blood cell count (RBC), hemoglobin (Hb), hematocrit (HCT), white blood cell count (WBC) and differential leukocyte counts including neutrophils (NE), lymphocytes (LY), monocytes (MO) and eosinophils (EO). NT-proBNP concentrations were measured using a fluorescence immunoassay analyzer (Vcheck V200, BIONOTE, Korea). Serum biochemical analysis, including C-reactive protein (CRP), was performed using a chemistry analyzer (Catalyst One, IDEXX Laboratories, Westbrook, ME, USA).
 
Radiographic examination
 
Thoracic radiographs were obtained at the time of presentation and again after complete resolution of clinical signs. Radiographic imaging was performed using an X-ray system (Accuray D5; DK, Seoul, Korea). Measurements of vertebral heart size (VHS) and vertebral left atrial size (VLAS) were conducted on right lateral thoracic radiographs obtained during inspiration. Radiographs were reviewed by two board-certified radiology specialists and VHS and VLAS were measured when pulmonary infiltrates had completely resolved (Fig 1).


 
Echocardiographic examination
 
Echocardiographic examinations were performed using an ultrasound system (Affiniti 50, Philips Medical Systems, Andover, MA, USA). The left atrium-to-aorta ratio (LA/Ao), left ventricular internal diameter in diastole normalized for body weight (LVIDdN) and early diastolic transmitral flow velocity (E-wave peak velocity) were measured to diagnose cardiogenic pulmonary edema (Fig 2). The diagnosis of Stage C MMVD was supported by left atrial enlargement (LA/Ao > 1.6), increased left ventricular internal diameter in diastole normalized for body weight (LVIDdN >1.7) and elevated transmitral E-wave velocity (>1.2 m/s), in accordance with the ACVIM consensus guidelines.


 
Furosemide treatment protocol
 
All dogs initially received an intravenous bolus injection of furosemide (4 mg/kg), followed by continuous rate infusion (CRI) at an initial rate of 0.7 mg/kg/h. The CRI rate was adjusted according to the patient’s respiratory rate, respiratory effort, hydration status and renal function. Continuous rate infusion was maintained until sufficient clinical improvement was achieved. Complete radiographic resolution of pulmonary edema was confirmed on follow-up thoracic radiographs.
 
Supportive treatment
 
During the acute management of cardiogenic pulmonary edema, oxygen supplementation was provided using flow-by oxygen or a face mask as clinically indicated. Oral pimobendan (0.25 mg/kg PO q12h) was administered whenever oral medication was clinically feasible. Intravenous fluid therapy was not routinely administered during continuous rate infusion of furosemide because of the risk of exacerbating pulmonary edema.
 
Statistical analysis
 
All collected data were coded and checked for errors before being compiled into an analysis dataset using Microsoft Excel. Statistical analyses were performed using IBM SPSS Statistics version 25.0 (SPSS Inc., Chicago, IL, USA). The level of statistical significance was set at p<0.05. Comparisons between groups based on prior heart medication use were conducted using the Mann-Whitney U test and pre- and post-treatment changes were analyzed using the Wilcoxon signed-rank test. Associations among body weight, vertebral heart size (VHS) and vertebral left atrial size (VLAS), including their change rates, were evaluated using Spearman’s rank correlation coefficient.

A total of 48 Maltese dogs with cardiogenic pulmonary edema were included in this study. Among these, 25 dogs had been newly diagnosed with cardiac disease at presentation, whereas 23 dogs had received cardiac medications prior to admission, including pimobendan (0.25 mg/kg, PO; Vetmedin 5 mg chewable tablets, Boehringer Ingelheim Vetmedica, Inc., Duluth, GA, USA) and enalapril (0.5 mg/kg, PO; Samnam enalapril, Samnam Pharmaceutical, Korea) (Table 1).


       
In dogs without prior cardiac medication use (Group 1), the sex distribution at the first episode of pulmonary edema was as follows: 13 castrated males (52%), 8 spayed females (32%) and 4 intact males (16%). In dogs receiving cardiac medications prior to admission (Group 2), the sex distribution consisted of 11 castrated males (47%), 7 spayed females (30%), 3 intact females (13%) and 2 intact males (8%).
       
The mean age ± standard deviation was 11.36 ± 2.27 years in Group 1 and 12.00±2.25 years in Group 2.
       
The mean body weight of all dogs at presentation was 3.56±1.30 kg, which decreased to 3.31±1.19 kg after complete resolution of pulmonary edema. In Group 1, body weight decreased from 3.60±1.23 kg to 3.34±1.14 kg, whereas in Group 2, it decreased from 3.53±1.40 kg to 3.28±1.28 kg (Table 2).


       
The mean vertebral heart size (VHS) of all dogs at presentation was 11.76±0.93, which decreased to 11.03±0.96 after resolution of pulmonary edema (Table 2). In Group 1, VHS decreased from 11.27±0.74 to 10.55±0.82, while in Group 2, it decreased from 12.30±0.82 to 11.56±0.83.
       
The mean vertebral left atrial size (VLAS) of all dogs at presentation was 3.04 ± 0.33, which decreased to 2.86±0.32 following resolution of pulmonary edema (Table 2). In Group 1, VLAS decreased from 2.92 ± 0.33 to 2.72 ± 0.32, whereas in Group 2, it decreased from 3.17 ± 0.27 to 3.00 ± 0.26.
       
The mean change rates of body weight, vertebral heart size (VHS) and vertebral left atrial size (VLAS) for all dogs are summarized in Table 3. The mean body weight loss rate for all dogs was -6.89 ± 2.40%. In Group 1, the mean body weight loss rate was -7.00 ± 2.66%, whereas in Group 2, it was -6.76±2.13%. The mean VHS change rate for all dogs was -6.19±4.16%. In Group 1, the mean VHS change rate was -6.40±4.20%, while in Group 2, it was -5.94±4.18%.


       
The mean VLAS change rate for all dogs was -5.97 ±4.36%. In Group 1, the mean VLAS change rate was -6.73 ±4.48%, whereas in Group 2, it was -5.13 ± 4.15%.
 
Correlation analysis by group
 
Correlations in dogs without prior heart medication (Group 1)
 
In Group 1, body weight at presentation (D0) showed a strong positive correlation with body weight after resolution of pulmonary edema (D1) (r = 0.997, p<0.01) and was also positively correlated with D0 vertebral heart size (VHS) (r = 0.521, p<0.01). Body weight at D1 was positively correlated with D0 VHS (r=0.535, p<0.01).
       
D0 VHS showed a strong positive correlation with D1 VHS (r=0.818, p<0.01) and was positively correlated with both D0 vertebral left atrial size (VLAS) and D1 VLAS (r=0.682 and 0.567, respectively; p<0.01). D0 VLAS was strongly positively correlated with D1 VLAS (r=0.913, p<0.01) (Table 4).


 
Correlations in dogs with prior heart medication (Group 2)
 
In Group 2, body weight at D0 showed a strong positive correlation with body weight at D1 (r=0.998, p<0.01). D0 VHS was positively correlated with D1 VHS (r=0.798, p<0.01).
       
D0 VLAS was positively correlated with both D1 VHS (r=0.512, p<0.05) and D1 VLAS (r=0.546, p<0.01). D1 VHS showed a positive correlation with VHS change rate (r=0.414, p<0.05) and was also positively correlated with both D0 VLAS and D1 VLAS (r=0.542 and 0.546, respectively; p<0.01). D0 VLAS was strongly positively correlated with D1 VLAS (r=0.874, p<0.01) (Table 5).


       
To further evaluate the study hypothesis, correlation analyses were performed using the entire study population (n = 48). The body weight change rate was not significantly correlated with either the VHS change rate (Spearman’s r = 0.027, p = 0.856) or the VLAS change rate (r = 0.246, p = 0.093) (Table 6).


 
Comparison between Group 1 and Group 2
 
There was no significant difference in age between dogs with and without prior heart medication use. The mean age was 11.36±2.27 years in Group 1 and 12.00±2.26 years in Group 2 (p = 0.333).
       
Regarding body weight loss rate, Group 1 showed a mean change of -7.01±2.66%, whereas Group 2 showed a mean change of -6.76±2.14%. Although the mean body weight loss rate was numerically greater in Group 1, the difference between groups was not statistically significant (p=0.726).
       
For the VHS change rate, the mean value was -6.41± 4.21% in Group 1 and -5.95 ± 4.18% in Group 2. Although Group 2 showed a slightly smaller reduction, no statistically significant difference was observed between the two groups (p = 0.706).
       
Similarly, the VLAS change rate was -6.73±4.48% in Group 1 and -5.13±4.15% in Group 2. Despite a numerically greater reduction in Group 1, the difference between groups was not statistically significant (p = 0.207) (Table 7).


       
In this retrospective study of 48 Maltese dogs diagnosed with cardiogenic pulmonary edema (CPE), the mean body weight at the time of CPE onset was 3.56 kg, which decreased to 3.31 kg after diuretic treatment and complete resolution of CPE, corresponding to a mean reduction of 6.89%. Previous studies have reported body weight values for dogs with MMVD at different ACVIM stages; however, longitudinal body weight changes within the same individuals have not been evaluated. Although weight loss associated with progression to ACVIM Stage C has been described, no studies have focused on the magnitude of weight reduction following resolution of CPE (Boswood et al., 2020). In the present study, complete radiographic resolution of cardiogenic pulmonary edema was observed with a mean body weight reduction of 6.89±2.40% (Table 3). This reduction most likely reflects effective intravascular fluid removal achieved during diuretic therapy rather than a true reduction in cardiac size. VHS is a widely used radiographic parameter for assessing cardiac enlargement and multiple studies have evaluated its change rate in dogs with heart disease (Franchini et al., 2021; Glaus et al., 2010). Previous studies have demonstrated that VHS increases with progression of MMVD and reflects progressive cardiac enlargement across different disease stages and breeds (Boswood et al., 2020; Mikawa et al., 2020). Recent studies have demonstrated that radiographic assessment remains valuable for monitoring improvement of pulmonary congestion in dogs with acute cardiogenic pulmonary edema (Koster et al., 2023). Similarly, the present study demonstrated a significant mean decrease in VHS of 6.19 ±4.16% after treatment. The observed reduction in VHS most likely reflects decreased cardiac filling pressure rather than a true reduction of cardiac dimensions.
       
VLAS is a radiographic index used to assess left atrial enlargement and has been increasingly studied in dogs with cardiac disease. Previous studies have shown that VLAS increases with progression of MMVD and correlates with left atrial enlargement in affected dogs (Mikawa et al., 2020; Lee et al., 2022). In the present study, VLAS decreased significantly by 5.97±4.36% after treatment. Similarly, the reduction in VLAS probably represents decreased left atrial distension secondary to reduced preload.
       
When comparing dogs receiving cardiac medications with those not receiving such medications, the change rates of VHS and VLAS were numerically greater in the medicated group; however, no statistically significant differences were observed. Although previous studies have demonstrated that pimobendan delays progression to congestive heart failure and improves clinical outcomes (Boswood et al., 2016; Boswood et al., 2018), few studies have specifically evaluated the effects of prior cardiac medication on changes in body weight and radiographic cardiac indices during treatment of cardiogenic pulmonary edema. Further studies are required to clarify the effects of cardiac medications on treatment outcomes in dogs with CPE.
       
These findings indicate that body weight loss and radiographic cardiac indices reflect different physiological processes during treatment. Therefore, body weight reduction should not be considered a direct surrogate marker for radiographic improvement. However, because body weight primarily reflects acute fluid removal whereas VHS and VLAS are influenced by cardiac geometry and loading conditions, body weight may still serve as a practical adjunctive clinical monitoring parameter, particularly when repeated thoracic radiography is not feasible in clinical practice.
       
This study has several limitations. First, only dogs with ACVIM Stage C CPE were included and prior echocardiographic and radiographic data were unavailable. Second, although body weight was measured at presentation and after resolution of CPE, urinary and fecal outputs were not standardized. Third, radiographic examinations and body weight measurements were performed based on urinary timing; therefore, complete resolution of CPE may have occurred before these assessments. Furthermore, standardized clinical variables such as respiratory rate, oxygen supplementation requirements and respiratory effort scores were not consistently available because of the retrospective study design. In addition, because the exact time of CRI discontinuation was not consistently recorded, the cumulative furosemide dose could not be reliably calculated. Previous studies have reported that cumulative parenteral furosemide administration may influence renal outcomes in dogs with left-sided congestive heart failure (Giorgi et al., 2022). Future prospective studies incorporating standardized clinical improvement indices and cumulative diuretic dose are warranted to determine whether body weight reduction can predict treatment response more accurately.

This retrospective study evaluated 48 Maltese dogs treated exclusively for CPE among 118 dogs receiving furosemide CRI between 2020 and 2022. Changes in body weight, VHS and VLAS and their associations with prior cardiac medication use were analyzed. Following furosemide CRI until radiographic resolution of pulmonary edema, mean body weight decreased by 6.89±2.40%, while VHS and VLAS decreased by 6.19±4.16% and 5.97±4.36%, respectively. No significant differences in these changes were observed between dogs with or without prior cardiac medication. Although limited by its retrospective design, this study demonstrates that body weight reduction consistently accompanies successful treatment of cardiogenic pulmonary edema. However, because body weight loss was not significantly associated with changes in radiographic cardiac indices, it should be regarded as an adjunctive clinical monitoring parameter rather than a surrogate marker of radiographic improvement.

The authors declare that this research was conducted independently and received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
 
Disclaimers
 
The views and conclusions expressed in this article are solely those of the authors and do not necessarily represent the views of their affiliated institutions. The authors are responsible for the accuracy and completeness of the information provided, but do not accept any liability for any direct or indirect losses resulting from the use of this content.
 
Informed consent
 
This study was a retrospective analysis of anonymized medical records obtained during routine clinical practice at a private veterinary hospital. Formal ethical approval was not required because no experimental interventions were performed. Written informed consent for the use of anonymized clinical records for research purposes was obtained from all owners at the time of admission.

The authors declare that there are no conflicts of interest regarding the publication of this article. No funding or sponsorship influenced the design of the study, data collection, analysis, decision to publish, or preparation of the manuscript.