Retrospective study pooling data from 4 previous samples including dogs with MMVD stage B2 or C. MRSI was calculated as: (heart rate [HR]/120) × left atrium-to-aorta ratio (LA:Ao) × (age in years/10) × 100. Alternative MRSI formulas substituting radiographic measures of left atrial size were also calculated. Cox proportional hazard modeling and time-dependent receiver-operator characteristic curves quantified prognostic performance.

Results

For Stage B2 pooled samples, MRSI > 156 was predictive of time to CHF (median 407 vs 1404 days; area under the curve [AUC] 0.68; hazard ratio 3.02 [95% CI 1.9-4.9]; P < .001). MRSI > 173 was predictive of all-cause death (median survival 868 vs 1843 days; AUC 0.64; hazard ratio 4.26 [95% CI 2.4-7.5]; P < .001). MRSI showed superior predictive value compared to the individual variables of HR, LA:Ao, and age. Variations of the MRSI equation substituting radiographic vertebral left atrial size for LA:Ao were also significantly predictive of outcome in stage B2. MRSI was not consistently predictive of outcome in Stage C.

Conclusions and Clinical Importance

MRSI was predictive of outcome (onset of CHF and all-cause death) in MMVD Stage B2, demonstrating utility as a useful prognostic tool. Echocardiographic LA:Ao can be effectively replaced by radiographically determined LA size in the MRSI formula.

Abbreviations

ACEi: angiotensin converting enzyme inhibitor
ACVIM: American College of Veterinary Internal Medicine
AUC: area under the curve
CHF: congestive heart failure
ECG: electrocardiogram
E-vel: E-wave transmitral peak velocity
FS%: left ventricular fractional shortening
HR: resting heart rate
LA:Ao: echocardiographic left atrium-to-aorta ratio
LVIDd: left ventricular internal diameter in diastole
LVIDs: left ventricular internal diameter in systole
MINE score: mitral insufficiency echocardiographic score
MMVD: myxomatous mitral valve disease
MRSI: mitral regurgitation severity index
MRSI-MVLAS: mitral regurgitation severity index utilizing modified vertebral left atrial size in place of echocardiographic left atrium-to-aorta ratio
MRSI-VHS: mitral regurgitation severity index utilizing vertebral heart size in place of echocardiographic left atrium-to-aorta ratio
MRSI-VLAS: mitral regurgitation severity index utilizing vertebral left atrial size in place of echocardiographic left atrium-to-aorta ratio
MST: median survival time
MVLAS: modified vertebral left atrial size
S (1-4): samples (1-4)
VHS: vertebral heart size
VLAS: vertebral left atrial size

1 INTRODUCTION

Myxomatous mitral valve disease (MMVD) is the most commonly diagnosed heart disease in dogs^{1, 2} and is variably progressive,^{3, 4} with some dogs experiencing cardiac-related death and others never progressing to congestive heart failure (CHF). Even after cardiac enlargement has developed (American College of Veterinary Internal Medicine [ACVIM] Stage B2),² the rate of progression is variable.^2-4 In addition, although most dogs die within 1 year of first-onset CHF secondary to MMVD,⁵ there is still relevant variability in survival time among individual dogs. This presents a challenge to clinicians when setting owner expectations, advising recheck scheduling, and determining appropriate aggressiveness of treatment.

Previous studies found associations between outcome in MMVD and clinical, echocardiographic, and radiologic variables. Variables with prognostic value include resting heart rate (HR),^{3, 5-9} echocardiographic left atrium-to-aorta ratio (LA:Ao),^{3-7, 10-12} left ventricular internal diameter in diastole (LVIDd)^{5, 7, 13} and systole (LVIDs),^{5, 14} transmitral peak velocity in early diastole (E-vel),^{3, 4, 6, 11, 12} left ventricular fractional shortening (FS%),⁷ vertebral heart size (VHS),^{5-7, 10} N-terminal pro-B-type natriuretic peptide^{6, 8-10, 13} and age.^{3, 9, 15} Previous investigators proposed indices combining several of these variables, including the PREDICT model,¹⁰ Mitral INsufficiency Echocardiographic (MINE) score,¹⁶ and Clinical Severity Scoring System.¹⁷ While such indices have prognostic utility, they require complex formulas or advanced echocardiographic measurements, limiting utility in the general practice setting.

The mitral regurgitation severity index (MRSI),¹⁸ defined as (HR/120) × LA:Ao × (age/10) × 100, was developed by the primary author of VETPROOF¹⁹ as a clinical prognostic index for the prediction of outcome in dogs with MMVD. As originally described, MRSI is a simple formula using 2 clinical variables (age and HR) and 1 basic echocardiographic variable (LA:Ao). The index (MRSI) was associated with survival in 1 sample of dogs with MMVD Stage B2 or C.²⁰ The primary objective of this retrospective study was to determine if MRSI accurately predicts outcome in MMVD in heterogenous samples of dogs with Stage B2 or C MMVD. We hypothesized that MRSI would predict time to onset or recurrence of CHF and all-cause death in dogs with MMVD, and that MRSI would show greater predictive ability in Stage B2 compared to Stage C. A secondary objective was to determine if the prognostic utility of MRSI exceeds that of its individual component variables alone or the MINE score. Our final objective was to determine if VHS or radiographic measurements of vertebral left atrial size (VLAS) can be substituted for LA:Ao in the MRSI equation to provide predictive indices (MRSI-VHS and MRSI-VLAS) that would be more accessible to general practitioners.

2 MATERIALS AND METHODS

2.1 Study design and study samples

This was a retrospective study of dogs with MMVD from 4 previous study samples (S1,²⁰ S2, S3,¹⁹ S4²¹). Specific inclusion and exclusion criteria for each study have been previously published for S1, S3, and S4.^19-21 Unpublished data from S2 was previously collated retrospectively by 1 of the authors (Seunggon Lee). The dogs included in S2 were presented to North Carolina State Veterinary Hospital between January 2009 and December 2011 with echocardiographic evidence of MMVD, a characteristic left apical systolic murmur, and first-onset CHF characterized by pulmonary edema, and were subsequently treated with furosemide, pimobendan, and an angiotensin converting enzyme inhibitor (ACEi). Medical records for S1 and S2 patients were retrospectively evaluated for additional qualifying study visits and long-term follow-up beyond the previous studies' data collection periods. Datasets from previous prospective studies (S3 and S4) were acquired with no additional record review. The present retrospective study did not require approval from the institutional animal care and use committee.

2.2 Definition of study visits

For S1 and S2, data from 2 study visits, asymptomatic and in CHF (ACVIM Stage B2 and Stage C, respectively),² were recorded for each patient when available (Figure 1). The Stage B2 visit was defined as the first visit with echocardiographic evidence of cardiomegaly attributed to MMVD, as defined by LA:Ao ≥ 1.6 (measured in M-mode or 2D) or documentation of moderate or severe left atrial enlargement on the echocardiogram report, or VHS > 10.5 if echocardiogram report was not available. The Stage C visit was defined as the first visit within 35 days of first-onset CHF. For Stage C visits, status as compensated vs decompensated at time of visit was recorded when this information was available. Dogs were defined as decompensated if there were clinical signs of dyspnea or tachypnea at the time of the visit or evidence of pulmonary edema noted on radiographs. Dogs were defined as compensated if the patient was eupneic at the time of the visit with no pulmonary edema noted on radiographs (eg, dogs treated for CHF before referral). Dogs were included in the study if they had at least 1 qualifying visit (Stage B2, Stage C, or both) with the following minimum information recorded: date of birth, HR, and either echocardiographic LA:Ao or radiographs available for review from which VHS and VLAS could be measured.

Details are in the caption following the image — **FIGURE 1**
Open in figure viewer PowerPoint

CONSORT diagram describing the 4 study samples, including number of dogs with Stage B2 and Stage C myxomatous mitral valve disease, medication requirements or restrictions, and availability of radiographs for each study sample. ACEi, angiotensin converting enzyme inhibitor; n, number of dogs.

2.3 Variables collected

The following demographics were collected from the initial qualifying visit for all study samples: birthdate, breed, sex, and weight. For both visits (Stage B2 and C), the following information was recorded when available: visit date, HR (from physical exam when available; otherwise from electrocardiogram or echocardiogram), heart murmur intensity, LA:Ao (right parasternal short-axis 2D^{7, 22} when available; otherwise M-mode²²), LVIDd from M-mode in the right parasternal short-axis view,^{23, 24} E-vel,^{16, 25} FS%,^{7, 16, 23} VHS,²⁶ vertebral left atrial size (VLAS),²⁷ modified vertebral left atrial size (MVLAS),²⁸ and concurrent cardiac and non-cardiac medications prescribed. Radiographic VHS, VLAS, and MVLAS measurements for S1 and S2 were performed post hoc by a single author (Michelle Vereb). The right lateral radiograph was used unless the borders of relevant structures were obscured, in which case the left lateral radiograph was used. Vertebral heart size was measured as previously described, extending the horizontal line to the intersection of the caudal border of the left atrium with dorsal border of the caudal vena cava.²⁶ The VLAS was measured by drawing a line connecting the most dorsal point on the vertical VHS line to the most caudal point on the horizontal VHS line, as previously described.²⁷ MVLAS was calculated by drawing a second line from the most dorsal border of the left atrium toward the VLAS line, intersecting the original VLAS line at a perpendicular angle.²⁸ These lines were compared to the length of the patient's vertebral bodies, starting at the cranial endplate of T4 and extending caudally, to the nearest 0.5 mm.^26-28 The MINE score was calculated for S1 and S2 dogs when all echocardiographic components of the index were available.¹⁶

2.4 Calculation of MRSI

For the Stage B2 visit, MRSI, MRSI-VHS, MRSI-VLAS, and MRSI-MVLAS were calculated; formulas are detailed in Table 1. Absolute values for individual components of MRSI were adjusted as follows to produce final MRSI values in a convenient range of whole numbers. Per the originally proposed MRSI formula, age and HR were normalized to the approximate sample medians from the VETPROOF study (10 years and 120 bpm, respectively); LA:Ao was not further adjusted. Radiographic measurements were normalized based on cutoffs previously associated with Stage B2 disease or presence of clinically relevant LA enlargement (10.5 for VHS,⁷ 2.5 for VLAS,²⁹ and 3.4 for MVLAS).²⁸ Based on initial exploratory analyses, MRSI was also calculated without age for Stage C visits (Table 1). Long-term outcomes were recorded as the date of first-onset CHF, date of recurrent CHF, date of death, and cause of death. For dogs lost to follow-up, the last contact date was recorded and cases were right-censored from the respective outcome analyses.

TABLE 1. Mitral regurgitation severity index (MRSI) equations evaluated for ability to predict time to congestive heart failure or death in dogs with Stage B2 and Stage C myxomatous mitral valve disease.

MRSI	HR/120 × LA:Ao × Age/10 × 100
MRSI-VHS	HR/120 × VHS/10.5 × Age/10 × 100
MRSI-VLAS	HR/120 × VLAS/2.5 × Age/10 × 100
MRSI-MVLAS	HR/120 × MVLAS/3.4 × Age/10 × 100
MRSI without age	HR/120 × LA:Ao × 100

Abbreviations: Age, in years; HR, heart rate in beats per minute; LA:Ao, left atrium-to-aorta ratio; MVLAS, modified vertebral left atrial score; VHS, vertebral heart score; VLAS, vertebral left atrial score.

2.5 Statistical analysis

All data are presented as median (range). To investigate if the different samples could be pooled for analysis, a shared frailty model was used with different MRSI formulas as fixed effects and study sample as a random effect. Samples were pooled if the frailty (heterogeneity in the data) was low and if pooling analysis using median classes showed P-value > .05 between samples. Within each individual study sample (S1-S4) and for pooled samples, Kaplan-Meier curves, Cox proportional hazard modeling, and time-dependent receiver-operator characteristic (ROC) curves at 1000 days quantified prognostic performance of MRSI, MRSI-C, MRSI-VHS, MRSI-VLAS, MRSI-MVLAS, individual components of MRSI (HR, age, LA:Ao, VHS, VLAS, MVLAS), and the MINE score to predict time to onset of CHF (Stage B2 visit) or relapse of CHF (Stage C visit) and time to all-cause death (Stage B2 and Stage C visits). For analyses of time to CHF onset or relapse and time to all-cause death, dogs lost to follow-up and dogs with no death date were censored, respectively. The proportional hazard assumption was confirmed using Schoenfeld's residuals. Martingale's residuals were used to assess goodness of fit for different models.

Constituent variables and MINE score were dichotomized using the sample median, whereas the maximally selected rank statistics determined the optimal MRSI cutoffs.³⁰ Area under the ROC curve (AUC; with 95% confidence intervals), hazard ratios (with 95% confidence intervals), sensitivity, and specificity were used to describe prognostic performance of different variables. The inverse probability of censoring weighting was used to estimate AUCs and compare AUCs among different variables. Predictive power based on AUC was described as follows: none (AUC 0.5), poor (AUC 0.5-0.7), acceptable (0.7-0.8), excellent (0.8-0.9), or outstanding (>0.9).³¹

To assess whether MRSI retained its prognostic ability in the face of differing cardiac treatments, the prognostic performance of MRSI and the specified treatment were analyzed separately in both univariable and multivariable analyses using forward selection. For all analyses, P-values < .05 were considered significant.

3 RESULTS

3.1 Demographic and clinical data

A total of 912 dogs were evaluated from the original study samples, and 43 dogs were excluded (Figure 1). A total of 869 dogs met the inclusion criteria, with 237 dogs having qualifying Stage B2 visits and 670 dogs having qualifying Stage C visits (Figure 1). The most common breeds, including mixed breeds with the respective breed predominating, included chihuahua (96), Cavalier King Charles spaniel (91), other mixed breed dogs (83), Maltese (54), shih tzu (54), dachshund (43), poodle (42), Yorkshire terrier (37), cocker spaniel (29), Pomeranian (27), beagle (24), and miniature schnauzer (24). Average dog weight was 9.0 kg (range 1.7-70.5 kg). There were 373 castrated males, 396 spayed females, 76 intact males, and 24 intact females. The median age was 10.3 years (range 5-17 years) at the Stage B2 visit and 11 years (range 3.3-19.2 years) at the Stage C visit. For Stage C visits in S1 and S2, 19 dogs (18.8%) were classified as compensated and 82 dogs (81.2%) were classified as decompensated.

3.2 Stage B2

Results of MRSI calculations and individual components of MRSI for both Stage B2 and C visits are summarized in Table 2. Overall incidence of missing data for MRSI variables was 0.3% (LA:Ao was missing for 2 dogs in S1 and 1 dog in S2; there were no missing data for S3 or S4). In Stage B2, MRSI was significantly associated with outcome (both time to CHF and survival) in each individual study sample (S1, S2, S3). A shared frailty model for B2 visits in S1-S3 showed low inter-variability, and pooling analysis using median classes revealed P > .10 for all variables, demonstrating that these samples could be pooled for further analyses. When samples S1-S3 were pooled, MRSI > 156 was predictive of time to CHF (median 407 days [95% CI 336-539 days] for MRSI > 156 vs 1404 days [95% CI 864-1933 days] for MRSI ≤ 156; HR = 3.02 [95% CI 1.88-4.87]; P < .001); see Table 3 and Figure 2. Also in the pooled S1-S3 sample, MRSI > 173 was predictive of all-cause death (median survival 868 days [95% CI 586-1193 days] for MRSI > 173 vs 1843 days [95% CI 1282-1938 days] for MRSI ≤ 171; HR = 2.76 [95% CI 1.75-4.37]; P < .001); see Table 3 and Figure 2. Individual component variables of MRSI (HR, age, and LA:Ao) were also significantly predictive of time to CHF. The AUC was significantly lower for HR (P = .02) and age (P = .04) compared to MRSI; the AUC for LA:Ao did not differ significantly from that of MRSI (P = .12). The only MRSI component variable that was predictive of time to death was age, with an AUC not significantly different from MRSI for this outcome (P = .52). In study samples where echocardiographic variables were available (S1 and S2), MRSI had similar predictive value to the MINE score (Table 4), with AUC values not significantly different from MRSI for any study sample (P = .74 for S1, P = .63 for S2, and P = .88 for S3).

TABLE 2. Summary statistics for mitral regurgitation severity index (MRSI) calculations and individual MRSI components in 869 dogs with Stage B2 or Stage C myxomatous mitral valve disease from 4 combined study samples.

	N	Median (range)	N	Median (range)
	Stage B2		Stage C
Age	237	10.3 (5-17)	670	11.0 (3.3-19.2)
Heart rate	237	126 (60-200)	670	140 (56-260)
LA:Ao	234	1.84 (1.10-2.91)	644	2.21 (1.49-3.85)
VHS	80	11.3 (9.0-14.3)	655	11.9 (9.0-15.4)
VLAS	80	2.5 (1.5-4.1)	88	2.9 (2.0-4.9)
MVLAS	79	3.5 (2.3-5.8)	87	3.7 (2.7-5.3)
MRSI	234	200 (60-560)
MRSI without age			644	259 (107-583)
MRSI-VHS	80	114 (46-246)
MRSI-VLAS	79	104 (36-314)
MRSI-MVLAS	80	101 (41-288)

Note: Formulas used to calculate MRSI are shown in Table 1.
Abbreviations: Age, in years; heart rate, in beats per minute; LA:Ao, left atrium-to-aorta ratio; MVLAS, modified vertebral left atrial score; N, number of dogs; VHS, vertebral heart score; VLAS, vertebral left atrial score.

TABLE 3. Prognostic performance of the Mitral Regurgitation Severity Index (MRSI) and its individual component variables to predict time to congestive heart failure or all-cause death in 234 dogs with Stage B2 myxomatous mitral valve disease from 3 pooled study samples (S1, n = 45; S2, n = 56; S3, n = 133).

		Cutoff	N	Days to event	AUC (95% CI)	Hazard ratio (95% CI)	Sens \| Spec	P-value
Time to CHF	MRSI	≤156	24 (41)	1404	0.73 (0.64-0.82)	3.02 (1.88-4.87)	0.79 \| 0.53	<.001
	MRSI	>156	87 (82)	407	0.73 (0.64-0.82)	3.02 (1.88-4.87)	0.79 \| 0.53	<.001
	Heart rate	≤126	52 (69)	778	0.62 (0.51-0.72)	1.50 (1.03-2.20)	0.89 \| 0.28	.03
	Heart rate	>126	61 (55)	456	0.62 (0.51-0.72)	1.50 (1.03-2.20)	0.89 \| 0.28	.03
	LA:Ao	≤1.84	49 (68)	907	0.65 (0.55-0.75)	1.93 (1.32-2.81)	0.58 \| 0.69	<.001
	LA:Ao	>1.84	62 (55)	434	0.65 (0.55-0.75)	1.93 (1.32-2.81)	0.58 \| 0.69	<.001
	Age	≤10	54 (56)	895	0.65 (0.56-0.75)	1.74 (1.20-2.55)	0.54 \| 0.72	.004
	Age	>10	59 (68)	372	0.65 (0.56-0.75)	1.74 (1.20-2.55)	0.54 \| 0.72	.004
Time to death	MRSI	≤173	21 (58)	1843	0.70 (0.60-0.80)	4.26 (2.43-7.47)	0.83 \| 0.61	<.001
	MRSI	>173	64 (91)	868	0.70 (0.60-0.80)	4.26 (2.43-7.47)	0.83 \| 0.61	<.001
	Heart rate	≤126	48 (73)	1294	0.56 (0.45-0.67)	1.33 (0.86-2.06)	0.50 \| 0.63	.2
	Heart rate	>126	39 (77)	1001	0.56 (0.45-0.67)	1.33 (0.86-2.06)	0.50 \| 0.63	.2
	LA:Ao	≤1.84	40 (77)	1357	0.58 (0.47-0.69)	1.50 (0.97-2.30)	0.53 \| 0.63	.07
	LA:Ao	>1.84	45 (72)	932	0.58 (0.47-0.69)	1.50 (0.97-2.30)	0.53 \| 0.63	.07
	Age	≤10	32 (78)	1843	0.67 (0.58-0.77)	2.58 (1.62-4.11)	0.63 \| 0.61	<.001
	Age	>10	55 (72)	910	0.67 (0.58-0.77)	2.58 (1.62-4.11)	0.63 \| 0.61	<.001

Note: Significant P values (P < .05) are indicated in bold. Formula used to calculate MRSI is shown in Table 1.
Abbreviations: Age, in years; AUC, area under the receiver operator characteristic curve; CHF, congestive heart failure; CI, confidence interval; heart rate, in beats per minute; LA:Ao, left atrium to aortic ratio; N, number of dogs reaching event in each analysis (with number of dogs censored in parenthesis); Sens, sensitivity; Spec, specificity.

TABLE 4. Prognostic performance of the mitral regurgitation severity index (MRSI) and the Mitral INsufficiency Echocardiographic Score (MINE) to predict time to congestive heart failure or all-cause death in 234 dogs with Stage B2 myxomatous mitral valve disease from 3 individual study samples (S1, n = 45; S2, n = 56; S3, n = 133).

	Variable and study sample	Cutoff	AUC (95% CI)	Hazard ratio (95% CI)	Sens \| Spec	P-value
Time to CHF	MRSI (S1)	≤184 vs >184	0.81 (0.60-1.00)	6.57 (1.45-29.69)	0.84 \| 0.71	.01
	MINE (S1)	≤8 vs >8	0.81 (0.61-1.00)	3.07 (1.06-8.88)	0.70 \| 0.57	.04
	MRSI (S2)	≤152 vs >152	0.85 (0.66-1.03)	4.28 (1.44-12.73)	0.84 \| 0.67	.009
	MINE (S2)	≤7 vs >7	0.80 (0.60-1.00)	2.35 (0.92-5.99)	0.73 \| 0.67	.07
	MRSI (S3)	≤156 vs >156	0.69 (0.57-0.81)	2.46 (1.38-4.38)	0.76 \| 0.50	.002
Time to death	MRSI (S1)	≤184 vs >184	0.72 (0.50-0.94)	4.25 (1.63-11.11)	0.79 \| 0.73	.003
	MINE (S1)	≤8 vs >8	0.69 (0.47-0.90)	1.47 (0.62-3.48)	0.45 \| 0.64	.38
	MRSI (S2)	≤162 vs >162	0.77 (0.57-0.97)	4.63 (1.59-13.52)	0.92 \| 0.55	.005
	MINE (S2)	≤7 vs >7	0.76 (0.59-0.93)	2.09 (0.89-4.87)	0.69 \| 0.64	.09
	MRSI (S3)	≤173 vs >173	0.69 (0.53-0.85)	4.32 (1.58-11.84)	0.83 \| 0.56	.004

Note: Significant P-values (P < .05) are bolded. Formula used to calculate MRSI is shown in Table 1.
Abbreviations: AUC, area under the receiver operator characteristic curve; CHF, congestive heart failure; CI, confidence interval; S1-S3, sample 1-3; Sens, sensitivity; Spec, specificity.

3.3 Stage B2 using radiographic data

The number of dogs with radiographs available (and thus with results for VHS, VLAS, and MVLAS) for each study sample are shown in Figure 1. When substituting radiographic measurements for LA:Ao in pooled S1 and S2 analysis, MRSI-VHS, MRSI-VLAS, and MRSI-MVLAS were significant predictors of outcome (Table 5). Predictive power (based on AUC and hazard ratio) for MRSI-VLAS and MRSI-MVLAS was similar to that of echocardiographic MRSI. Radiographic measurements of LA size alone (VLAS and MVLAS) also had similar predictive power to MRSI equations incorporating these variables.

TABLE 5. Prognostic performance of the mitral regurgitation severity index (MRSI) for 234 dogs with Stage B2 myxomatous mitral valve disease from 3 pooled study samples (S1, n = 45; S2, n = 56; S3, n = 133) and prognostic performance of MRSI with substitution of radiographic measurements for echocardiographic LA:Ao (MRSI-VHS, MRSI-VLAS, MRSI-MVLAS) for 80 dogs with Stage B2 myxomatous mitral valve disease from 2 pooled study samples (S1, n = 35; S2, n = 45).

	Variable	Cutoff	Number of dogs	Days to event	AUC (95% CI)	Hazard ratio (95% CI)	Sens \| Spec	P-value
Time to CHF	MRSI	≤170	10 (25)	1404	0.79 (0.66-0.92)	3.02 (1.88-4.87)	0.83 \| 0.71	<.001
	MRSI	>170	32 (34)	301	0.79 (0.66-0.92)	3.02 (1.88-4.87)	0.83 \| 0.71	<.001
	MRSI-VHS	≤111	14 (22)	1259	0.74 (0.59-0.90)	2.42 (1.12-5.22)	0.66 \| 0.70	.02
	MRSI-VHS	>111	21 (23)	450	0.74 (0.59-0.90)	2.42 (1.12-5.22)	0.66 \| 0.70	.02
	MRSI-VLAS	≤123	19 (35)	1259	0.77 (0.62-0.93)	5.06 (2.36-10.88)	0.52 \| 1.0	<.001
	MRSI-VLAS	>123	16 (10)	225	0.77 (0.62-0.93)	5.06 (2.36-10.88)	0.52 \| 1.0	<.001
	MRSI-MVLAS	≤125	18 (34)	1259	0.79 (0.64-0.94)	5.29 (2.43-11.51)	0.55 \| 1.0	<.001
	MRSI-MVLAS	>125	17 (10)	225	0.79 (0.64-0.94)	5.29 (2.43-11.51)	0.55 \| 1.0	<.001
	VHS	≤11.3	16 (25)	825	0.57 (0.36-0.77)	1.74 (0.86-3.52)	0.52 \| 0.50	.12
	VHS	>11.3	19 (20)	333	0.57 (0.36-0.77)	1.74 (0.86-3.52)	0.52 \| 0.50	.12
	VLAS	≤2.5	15 (30)	1259	0.76 (0.60-0.93)	3.44 (1.68-7.07)	0.62 \| 0.80	<.001
	VLAS	>2.5	20 (15)	298	0.76 (0.60-0.93)	3.44 (1.68-7.07)	0.62 \| 0.80	<.001
	MVLAS	≤3.5	15 (30)	1259	0.75 (0.58-0.92)	3.08 (1.50-6.31)	0.62 \| 0.80	.002
	MVLAS	>3.5	20 (14)	301	0.75 (0.58-0.92)	3.08 (1.50-6.31)	0.62 \| 0.80	.002
Time to death	MRSI	≤169	16 (18)	1509	0.71 (0.58-0.85)	4.26 (2.43-7.47)	0.83 \| 0.76	<.001
	MRSI	>169	41 (26)	595	0.71 (0.58-0.85)	4.26 (2.43-7.47)	0.83 \| 0.76	<.001
	MRSI-VHS	≤93	10 (15)	1509	0.75 (0.61-0.90)	3.12 (1.45-6.72)	0.81 \| 0.65	.004
	MRSI-VHS	>93	25 (30)	673	0.75 (0.61-0.90)	3.12 (1.45-6.72)	0.81 \| 0.65	.004
	MRSI-VLAS	≤78	12 (12)	1843	0.76 (0.61-0.90)	4.16 (1.80-9.61)	0.84 \| 0.59	<.001
	MRSI-VLAS	>78	32 (24)	657	0.76 (0.61-0.90)	4.16 (1.80-9.61)	0.84 \| 0.59	<.001
	MRSI-MVLAS	≤84	14 (11)	1509	0.78 (0.64-0.92)	3.47 (1.61-7.48)	0.83 \| 0.71	.002
	MRSI-MVLAS	>84	29 (25)	657	0.78 (0.64-0.92)	3.47 (1.61-7.48)	0.83 \| 0.71	.002
	VHS	≤11.3	25 (16)	844	0.47 (0.30-0.64)	1.12 (0.61-2.06)	0.43 \| 0.47	.71
	VHS	>11.3	19 (20)	717	0.47 (0.30-0.64)	1.12 (0.61-2.06)	0.43 \| 0.47	.71
	VLAS	≤2.5	23 (22)	962	0.70 (0.55-0.85)	1.86 (1.01-3.40)	0.49 \| 0.71	.05
	VLAS	>2.5	21 (14)	564	0.70 (0.55-0.85)	1.86 (1.01-3.40)	0.49 \| 0.71	.05
	MVLAS	≤3.5	24 (21)	1282	0.69 (0.54-0.84)	2.51 (1.31-4.79)	0.51 \| 0.82	.005
	MVLAS	>3.5	19 (15)	564	0.69 (0.54-0.84)	2.51 (1.31-4.79)	0.51 \| 0.82	.005

Note: Significant P values (P < .05) are bolded. Formulas used to calculate MRSI are shown in Table 1. VLAS was unreadable for 1 dog in S1.
Abbreviations: AUC, area under the receiver operator characteristic curve; CHF, congestive heart failure; MVLAS, modified vertebral left atrial size; N, number of dogs reaching event in each analysis (number of dogs censored in parenthesis); Sens, sensitivity; Spec, specificity; VHS, vertebral heart size; VLAS, vertebral left atrial size.

3.4 Stage C

In initial analyses, MRSI was not a significant predictor of outcome in Stage C in S1, S2, or S4. In fact, older age was associated with longer time to recurrent CHF in S4 (P = .05), suggesting that age had the opposite impact on outcome in Stage C compared to B2, which would necessarily impact prognostic power of MRSI in this stage. When age was removed from MRSI calculation, MRSI without age was a significant predictor of both recurrent CHF and all-cause death in Stage C dogs in S4, but was not a significant predictor of either outcome in S1 or S2 (see Table S1). Stage C visit data from S1, S2, and S4 were not pooled because of the differing patterns of response among samples.

3.5 Effect of treatment

For S1 and S2, a total of 53 Stage B2 dogs (51%) received pimobendan and 51 Stage B2 dogs (49%) did not; all but 6 dogs received an ACEi. For S3, 67 Stage B2 dogs (50%) received enalapril whereas 66 Stage B2 dogs (50%) received placebo. For S4, 284 Stage C dogs (50%) received benazepril + spironolactone (CARDALIS, Ceva Santé Animale, Libourne, France) whereas 285 Stage C dogs (50%) received benazepril; all dogs received furosemide. In multivariable Cox regression models including MRSI, treatment (pimobendan, enalapril, benazepril, or benazepril + spironolactone), and the interaction between MRSI and treatment, no treatment analyzed had a significant effect on the ability of MRSI to predict outcomes (see Table S2).

4 DISCUSSION

The purpose of this study was to investigate potential prognostic utility of MRSI in a large and diverse sample of dogs with MMVD from multiple institutions. The MRSI, as well as indices replacing LA:Ao with radiographic measurements of left atrial size, significantly predicted onset of CHF and all-cause death in stage B2. However, in Stage C, MRSI was only associated with outcome in a single study sample after removal of age from the formula.

Our hypothesis regarding the prognostic value of MRSI was supported in Stage B2, as MRSI was predictive of time to first-onset CHF and all-cause death in both the individual sample analyses and pooled analysis. For Stage B2 pooled samples, MRSI significantly predicted both time to CHF and time to death with an approximately 1000 day difference between group medians. However, the predictive power of MRSI was modest, in the range defined as acceptable (AUC 0.7-0.8). While MRSI outperformed 2 of its component variables (HR and age), there was no significant difference between AUC values for MRSI and LA:Ao. In addition, MRSI performed similarly to the MINE score in these study samples. While the MINE score is a clinically useful composite echocardiographic index,¹⁶ MRSI gives clinicians an additional option for MMVD prognostication that does not require spectral Doppler measurements of mitral inflow.

When radiographic measurements of left atrial size (VLAS and MVLAS) were substituted for LA:Ao, MRSI-VLAS and MRSI-MVLAS performed similarly to echocardiographic MRSI for prediction of outcome in stage B2, with AUC values also in the acceptable range (0.7-0.8). Furthermore, VLAS and MVLAS alone performed similarly to MRSI equations containing these variables, particularly for predicting time to CHF. These results confirm the utility of VLAS as an indicator of left atrial size^{27-29, 32, 33} and suggest a prognostic role for VLAS and MVLAS in MMVD stage B2. While MRSI-VHS was predictive of outcome, VHS alone was not, consistent with a previous study showing that VLAS predicted echocardiographic left atrial enlargement more accurately than VHS.³³

Many of the candidate predictor variables and equations investigated this study performed similarly to predict outcome in MMVD stage B2, with predictive power in the range of poor to acceptable (AUC 0.6-0.8). These variables and equations might have utility in different clinical settings depending on what diagnostic testing is available. Our results suggest that VLAS, MVLAS, or MRSI equations using these variables are useful in situations where thoracic radiographs can be performed. The original MRSI equation might have the most utility in situations where simple echocardiographic measurements are available, but radiographs are not. Finally, the MINE score remains an additional option when more advanced echocardiography can be performed.

For the Stage C visit, age had an opposite effect than expected in initial analysis of S4 using MRSI, with younger dogs reaching the outcomes of recurrent CHF or death sooner than older dogs. It is possible that dogs diagnosed with CHF at an earlier age have more rapidly progressive disease, therefore shortening their survival time, whereas older dogs experiencing first-onset CHF might have more slowly progressive disease or superior ability to compensate. Even when age was removed from the equation, MRSI without age was only predictive of both CHF recurrence and death in S4. This could be because of the larger sample size in S4 compared to S1 or S2. Additionally, inclusion criteria were stricter for S4, potentially resulting in a more uniform sample. Larger studies with a longer follow-up period could be considered to further evaluate the prognostic utility MRSI in Stage C dogs.

Overall, MRSI showed better prognostic utility in Stage B2 when compared to Stage C, also supporting our hypothesis. Stage B2 is arguably more important to prognosticate since progression in stage B2 MMVD can be quite variable among dogs,^{3, 4} with some dogs never developing CHF and others experiencing first-onset CHF anywhere from months to several years after diagnosis of Stage B2. Prognosis in Stage C is less variable, with most dogs treated with heart failure medications experiencing cardiac-related death within 6-14 months of first-onset CHF.^{1, 5, 34, 35}

Although the purpose of this study was not to analyze effectiveness of specific medical therapies, analyses were performed to determine whether MRSI retained its prognostic ability regardless of cardiac treatment. As hypothesized, treatment with pimobendan (Stage B2 in S1 and S2), enalapril (Stage B2 in S3), or spironolactone (Stage C in S4) did not significantly affect the predictive power of MRSI. This is likely because of the dramatic dichotomization of Stage B2 dogs using the optimal MRSI cut-off, with an approximately 33-month difference in time to CHF between dogs with low vs high MRSI scores. Comparatively, the benefit of treatment in extending the preclinical period for dogs with Stage B2 MMVD is a shorter period of time, in the range of 15 months for pimobendan⁷ and 4 months for enalapril.¹⁹ It is therefore not surprising that the predictive power of MRSI was independent of cardiac treatment, since differences in outcome predicted by MRSI were substantially greater than differences in outcome expected from treatment.

There were several inherent limitations in this retrospective study. Multiple study samples with different inclusion criteria were pooled for these analyses. For example, S3 and S4 were originally recruited in the context of a prospective study, whereas S1 and S2 were not, making the inclusion criteria less stringent for S1 and S2 as compared to S3 and S4. Sample S3 included only dogs ≤20 kg and S4 included only dogs ≥2.5 kg, whereas there were no weight restrictions for S1 or S2. In addition, not all variables were collected for each dog and availability of medical records for review, such as original radiographs, varied between samples. In rare circumstances, echocardiographic reports did not include a quantitative measurement of LA:Ao, and inclusion and classification of dogs as Stage B2 was based on a subjective description in the report (eg, “severe LA dilatation”) in combination with other echocardiographic variables, VHS, and VLAS. Treatments were not standardized, and follow-up data was missing for some dogs. Heart rate was measured at different times during the visit and by operators with varying experience. Multiple operators performed and analyzed echocardiograms. Although VLAS and VHS were measured post hoc by a single operator and preference was given to the right lateral radiograph for consistency, if the relevant structural borders were obscured, the left lateral projection was used. While a previous study found an insignificant difference between VLAS obtained on the right vs left lateral projection, comparing measurements obtained on differing views is not advised because of the wide range of agreement.²⁷ This study assessed both VLAS and MVLAS, whereas other published methods of measuring LA size on thoracic radiographs were not evaluated in this study, including radiographic LA dimension³⁶ and LA width³⁷; like VLAS, the latter measurement does not require identification of the dorsal LA border radiographically. While the study benefitted from a large sample size, analyses were focused on dichotomization; if even larger samples were used for Stage B2, tertiles could possibly further refine predictive cutoffs. This study was designed as an initial exploration into whether MRSI could potentially provide clinically useful information; ideally, these results should be validated using an external (and larger) dataset not used in model development. Finally, MRSI components (including age, HR, and radiographic measurements) were adjusted by dividing a given dog's value by a standard denominator to produce final MRSI calculations in a convenient range of whole numbers. For age and HR, denominators were chosen as originally described based on approximate median values from the VETPROOF study. For radiographic measurements (VHS, VLAS, and MVLAS), we chose denominators corresponding to the cutoff values for clinically relevant cardiomegaly enlargement. The authors acknowledge that the choice of denominators was somewhat arbitrary, and calculated MRSI values would differ if other values were chosen. However, since the same denominators were applied consistently to MRSI calculations for all dogs with simple division, choice of denominator would not impact the prognostic performance of MRSI.

5 CONCLUSIONS

The MRSI is a simple, easily obtained, clinically useful prognostic index for MMVD Stage B2 in dogs. Echocardiographic MRSI might be particularly useful if simple echocardiographic measurements are available but thoracic radiographs are not. Radiographic estimates of left atrial size (VLAS and MVLAS) can be substituted into the MRSI equation if echocardiography is not available and can also be used alone. Prognostic predictors identified in this study could potentially be applied to both specialty and general practice to better inform client expectations, guide the aggressiveness of treatment, and direct frequency of recheck examinations. In a research setting, MRSI could help ensure equal disease severity within study groups. This study forms a basis for further refinements to this type of prognostic index. Further studies, including validation of the model with large external datasets and prospective investigations, are warranted to confirm the utility of MRSI in general and specialty practice.

ACKNOWLEDGMENT

No funding was received for this study. Presented in abstract form at the 2022 American College of Veterinary Internal Medicine (ACVIM) Forum, Austin, TX. Open access funding provided by the Iowa State University Library.

CONFLICT OF INTEREST DECLARATION

Drs. Blondel, Coffman, and Guillot are employees of Ceva Santé Animale/Ceva Animal Health. Drs. Atkins, Adin, and Ward have received consulting fees and honoraria from Ceva Santé Animale. As a veterinary pharmaceutical company, Ceva develops and markets cardiovascular medications for dogs with MMVD, and sponsored 1 of the studies for which a dataset was utilized in the present retrospective study. However, the purpose and results of the present retrospective study deal only with prognosis for MMVD, not with the impact or outcome of any specific drug. No other authors declare a conflict of interest.

OFF-LABEL ANTIMICROBIAL DECLARATION

Authors declare no off-label use of antimicrobials.

INSTITUTIONAL ANIMAL CARE AND USE COMMITTEE (IACUC) OR OTHER APPROVAL DECLARATION

Authors declare no IACUC or other approval was needed.

HUMAN ETHICS APPROVAL DECLARATION

Authors declare human ethics approval was not needed for this study.

Supporting Information

REFERENCES

1Borgarelli M, Buchanan JW. Historical review, epidemiology and natural history of degenerative mitral valve disease. J Vet Cardiol. 2012; 14: 93-101.
10.1016/j.jvc.2012.01.011
PubMed Web of Science® Google Scholar
2Keene BW, Atkins CE, Bonagura JD, et al. ACVIM consensus guidelines for the diagnosis and treatment of myxomatous mitral valve disease in dogs. J Vet Intern Med. 2019; 33: 1127-1140.
10.1111/jvim.15488
PubMed Web of Science® Google Scholar
3Borgarelli M, Savarino P, Crosara S, et al. Survival characteristics and prognostic variables of dogs with mitral regurgitation attributable to myxomatous valve disease. J Vet Intern Med. 2008; 22: 120-128.
10.1111/j.1939-1676.2007.0008.x
CAS PubMed Web of Science® Google Scholar
4Borgarelli M, Crosara S, Lamb K, et al. Survival characteristics and prognostic variables of dogs with preclinical chronic degenerative mitral valve disease attributable to myxomatous degeneration. J Vet Intern Med. 2012; 26: 69-75.
10.1111/j.1939-1676.2011.00860.x
CAS PubMed Web of Science® Google Scholar
5Häggström J, Boswood A, O'Grady M, et al. Effect of pimobendan or benazepril hydrochloride on survival times in dogs with congestive heart failure caused by naturally occurring myxomatous mitral valve disease: the QUEST study. J Vet Intern Med. 2008; 22: 1124-1135.
10.1111/j.1939-1676.2008.0150.x
CAS PubMed Web of Science® Google Scholar
6Borgarelli M, Ferasin L, Lamb K, et al. DELay of appearance of symptoms of canine degenerative mitral valve disease treated with spironolactone and benazepril: the DELAY study. J Vet Cardiol. 2020; 27: 34-53.
10.1016/j.jvc.2019.12.002
CAS PubMed Web of Science® Google Scholar
7Boswood A, Häggström J, Gordon SG, et al. Effect of pimobendan in dogs with preclinical myxomatous mitral valve disease and cardiomegaly: the EPIC study – a randomized clinical trial. J Vet Intern Med. 2016; 30: 1765-1779.
10.1111/jvim.14586
CAS PubMed Web of Science® Google Scholar
8Mattin MJ, Brodbelt DC, Church DB, Boswood A. Factors associated with disease progression in dogs with presumed preclinical degenerative mitral valve disease attending primary care veterinary practices in the United Kingdom. J Vet Intern Med. 2018; 33: 445-454.
10.1111/jvim.15390
PubMed Web of Science® Google Scholar
9Mattin MJ, Boswood A, Church DB, Brodbelt DC. Prognostic factors in dogs with presumed degenerative mitral valve disease attending primary-care veterinary practices in the United Kingdom. J Vet Intern Med. 2019; 33: 432-444.
10.1111/jvim.15251
CAS PubMed Web of Science® Google Scholar
10Reynolds CA, Brown DC, Rush JE, et al. Prediction of first onset of congestive heart failure in dogs with degenerative mitral valve disease: the PREDICT cohort study. J Vet Cardiol. 2012; 14: 193-202.
10.1016/j.jvc.2012.01.008
PubMed Web of Science® Google Scholar
11Sargent J, Muzzi R, Mukherjee R, et al. Echocardiographic predictors of survival in dogs with myxomatous mitral valve disease. J Vet Cardiol. 2015; 17: 1-12.
10.1016/j.jvc.2014.11.001
PubMed Web of Science® Google Scholar
12Nakamura K, Osuga T, Morishita K, et al. Prognostic value of left atrial function in dogs with chronic mitral valvular heart disease. J Vet Intern Med. 2014; 28: 1746-1752.
10.1111/jvim.12441
CAS PubMed Web of Science® Google Scholar
13Moonarmart W, Boswood A, Luis Fuentes V, et al. N-terminal pro B-type natriuretic peptide and left ventricular diameter independently predict mortality in dogs with mitral valve disease. J Small Anim Pract. 2010; 51: 84-96.
10.1111/j.1748-5827.2009.00889.x
CAS PubMed Web of Science® Google Scholar
14Reimann MJ, Møller JE, Häggström J, et al. Mitral regurgitation severity and left ventricular systolic dimension predict survival in young Cavalier King Charles spaniels. J Vet Intern Med. 2017; 31: 1008-1016.
10.1111/jvim.14759
CAS PubMed Web of Science® Google Scholar
15Kvart C, Häggström J, Pedersen HD, et al. Efficacy of enalapril for prevention of congestive heart failure in dogs with myxomatous valve disease and asymptomatic mitral regurgitation. J Vet Intern Med. 2002; 16: 80-88.
10.1111/j.1939-1676.2002.tb01610.x
CAS PubMed Web of Science® Google Scholar
16Vezzosi T, Grosso G, Tognetti R, et al. The Mitral INsufficiency Echocardiographic score: a severity classification of myxomatous mitral valve disease in dogs. J Vet Intern Med. 2021; 35: 1238-1244.
10.1111/jvim.16131
PubMed Web of Science® Google Scholar
17López-Alvarez J, Elliott J, Pfeiffer D, et al. Clinical severity score system in dogs with degenerative mitral valve disease. J Vet Intern Med. 2015; 29: 575-581.
10.1111/jvim.12544
CAS PubMed Web of Science® Google Scholar
18Atkins C, Adin D, Blondel T, et al. Mitral regurgitation severity index: a simple calculation for classification and prognostication? Presented at: ACVIM Forum 2022; 23 Jun 2022, Austin, TX.
Google Scholar
19Atkins CE, Keene BW, Brown WA, et al. Results of the veterinary enalapril trial to prove reduction in onset of heart failure in dogs chronically treated with enalapril alone for compensated, naturally occurring mitral valve insufficiency. J Am Vet Med Assoc. 2007; 231: 1061-1069.
10.2460/javma.231.7.1061
CAS PubMed Web of Science® Google Scholar
20Ward JL, Chou YY, Yuan L, Dorman KS, Mochel JP. Retrospective evaluation of a dose-dependent effect of angiotensin-converting enzyme inhibitors on long-term outcome in dogs with cardiac disease. J Vet Intern Med. 2021; 35: 2102-2111.
10.1111/jvim.16236
PubMed Web of Science® Google Scholar
21Coffman M, Guillot E, Blondel T, et al. Clinical efficacy of a benazepril and spironolactone combination in dogs with congestive heart failure due to myxomatous mitral valve disease: the BEnazepril Spironolactone STudy (BESST). J Vet Intern Med. 2021; 35: 1673-1687.
10.1111/jvim.16155
PubMed Web of Science® Google Scholar
22Hansson K, Häggström J, Kvart C, Lord P. Left atrial to aortic root indices using two-dimensional and M-mode echocardiography in cavalier King Charles spaniels with and without left atrial enlargement. Vet Radiol Ultrasound. 2002; 43: 568-575.
10.1111/j.1740-8261.2002.tb01051.x
CAS PubMed Web of Science® Google Scholar
23Cornell CC, Kittleson MD, Della Torre P, et al. Allometric scaling of M-mode cardiac measurements in normal adult dogs. J Vet Intern Med. 2004; 18: 311-321.
10.1111/j.1939-1676.2004.tb02551.x
PubMed Web of Science® Google Scholar
24Thomas WP, Gaber CE, Jacobs GJ, et al. Recommendations for standards in transthoracic two-dimensional echocardiography in the dog and cat. J Vet Intern Med. 1993; 7: 247-252.
10.1111/j.1939-1676.1993.tb01015.x
CAS PubMed Web of Science® Google Scholar
25Schober K, Hart TM, Stern JA, et al. Detection of congestive heart failure in dogs by Doppler echocardiography. J Vet Intern Med. 2010; 24: 1358-1368.
10.1111/j.1939-1676.2010.0592.x
CAS PubMed Web of Science® Google Scholar
26Buchanan JW. Vertebral scale system to measure heart size in radiographs. Vet Clin North Am Small Anim Pract. 2000; 30: 379-393.
10.1016/S0195-5616(00)50027-8
CAS PubMed Web of Science® Google Scholar
27Malcolm EL, Visser LC, Phillips KL, Johnson LR. Diagnostic value of vertebral left atrial size as determined from thoracic radiographs for assessment of left atrial size in dogs with myxomatous mitral valve disease. J Vet Intern Med. 2018; 253: 1038-1045.
Google Scholar
28Lam C, Gavaghan BJ, Meyers FE. Radiographic quantification of left atrial size in dogs with myxomatous mitral valve disease. J Vet Intern Med. 2021; 35: 747-754.
10.1111/jvim.16073
PubMed Web of Science® Google Scholar
29Mikawa S, Nagakawa M, Ogi H, et al. Use of vertebral left atrial size for staging of dogs with myxomatous valve disease. J Vet Cardiol. 2020; 30: 92-99.
10.1016/j.jvc.2020.06.001
CAS PubMed Web of Science® Google Scholar
30Lausen B, Schumacher M. Maximally selected rank statistics. Biometrics. 1992; 48: 73-85.
10.2307/2532740
Web of Science® Google Scholar
31Hosmer DW, Lemeshow S, Sturdivant RX. Applied Logistic Regression. 3rd ed. Hoboken, NJ: John Wiley & Sons, Ltd; 2013.
10.1002/9781118548387
Web of Science® Google Scholar
32Vezzosi T, Puccinelli C, Tognetti R, Pelligra T, Citi S. Radiographic vertebral left atrial size: a reference interval study in healthy adult dogs. Vet Radiol Ultrasound. 2020; 61: 507-511.
10.1111/vru.12896
PubMed Web of Science® Google Scholar
33Duler L, Visser LC, Jackson KN, Phillips KL, Pollard RE, Wanamaker MW. Evaluation of radiographic predictors of left heart enlargement in dogs with known or suspected cardiovascular disease. Vet Radiol Ultrasound. 2021; 62: 271-281.
10.1111/vru.12949
PubMed Web of Science® Google Scholar
34Ettinger S, Benitz A, Ericsson G, et al. Effects of enalapril maleate on survival of dogs with naturally acquired heart failure. The Long-Term Investigation of Veterinary Enalapril (LIVE) Study Group. J Am Vet Med Assoc. 1998; 213: 1573-1577.
10.2460/javma.1998.213.11.1573
CAS PubMed Web of Science® Google Scholar
35 BENCH (BENazepril in Canine Heart disease) Study Group. The effect of benazepril on survival times and clinical signs of dogs with congestive heart failure: results of a multicenter, prospective, randomized, double-blinded, placebo-controlled, long-term clinical trial. J Vet Cardiol. 1999; 1: 7-18.
10.1016/S1760-2734(06)70025-X
Google Scholar
36Levicar C, Granados-Soler JL, Freise F, Raue JF, Nolte I, Bach JP. Comparison of different radiographic scores with associated echocardiographic measurements and prediction of heart enlargement in dogs with and without myxomatous mitral valve disease. J Vet Cardiol. 2022; 44: 1-12.
10.1016/j.jvc.2022.08.004
CAS PubMed Web of Science® Google Scholar
37Stepien RL, Rak MB, Blume LM. Use of radiographic measurements to diagnose stage B2 preclinical myxomatous mitral valve disease in dogs. J Am Vet Med Assoc. 2020; 256: 1129-1136.
10.2460/javma.256.10.1129
CAS PubMed Web of Science® Google Scholar

Citing Literature

Volume38, Issue1

January/February 2024

Pages 51-60

Efficacy of a mitral regurgitation severity index to predict long-term outcome in dogs with myxomatous mitral valve disease