Abstract
Objectives
Large studies focusing on restrictive cardiomyopathy (RCM) in the cat are scarce. The aims of this retrospective study were to describe epidemiological characteristics and to analyse prognostic factors affecting survival in cats with RCM.
Methods
The clinical archives of the Gran Sasso Veterinary Clinic (Milan, Italy) and of the cardiology unit of the Department of Veterinary Medicine (University of Milan, Italy) from 1997–2015 were reviewed for all cats diagnosed with RCM based on an echocardiographic examination (left atrial/biatrial enlargement, normal left ventricle wall thickness, normal or mildly decreased systolic function and restrictive left ventricle filling pattern with pulsed Doppler echocardiography).
Results
The study population comprised 90 cats (53 male, 37 female) with an echocardiographic diagnosis of RCM. Most were domestic shorthairs (n = 60) with a mean ± SD age of 10.0 ± 4.3 years and a median weight of 3.8 kg (interquartile range 3.2–5 kg). Most cats were symptomatic (n = 87). The most common clinical sign was respiratory distress (n = 75). Follow-up was available for 60 cats and the median survival time (MST) was 69 days (95% confidence interval [CI] 0–175 days). Cardiac-related death occurred in 50 cats. In the multivariate Cox analysis only respiratory distress showed a statistically significant effect on survival. The cats without respiratory distress showed an MST of 466 days (95% CI 0–1208); cats with respiratory distress showing an MST of 64 days (95% CI 8–120; P = 0.011).
Conclusions and relevance
RCM can be considered an end-stage condition associated with a poor prognosis, with few cats not showing clinical signs and surviving >1 year. Most cats died of cardiac disease within a very short time.
Introduction
Restrictive cardiomyopathy (RCM) is a myocardial disorder characterised by myocardial stiffness, severe diastolic dysfunction (restrictive physiology) and an overall poor prognosis.1–4 It is not clear if some of the cases of RCM should be considered the end result of other forms of cardiomyopathy (CM), mainly hypertrophic cardiomyopathy (HCM) and myocarditis.2,5–10 As serial echocardiographic examinations are seldom available for review to substantiate changes in the echocardiographic appearance, it is difficult to quantify or identify whether different separate aetiologies contribute to a common end-stage pattern. RCM is often morphologically subclassified into two forms: myocardial and endomyocardial. 1 Echocardiography allows classification and is currently the most common tool for diagnosis. Both forms of RCM are characterised by atrial enlargement, normal left ventricular (LV) wall thickness, normal or mildly decreased systolic function, and restrictive LV filling pattern with pulsed Doppler echocardiography; in the endomyocardial form, thick hyperechoic tissues bridge the LV lumen.1–4,8 Large studies focusing on RCM in the cat are scant.2,3 The aims of this retrospective study were to describe epidemiological characteristics and to analyse prognostic factors affecting survival in cats with RCM.
Materials and methods
The clinical archives of the Gran Sasso Veterinary Clinic (Milan, Italy) and of the cardiology unit of the Department of Veterinary Medicine (University of Milan, Italy) from 1997–2015 were reviewed to identify cats diagnosed with RCM based on an echocardiographic examination. Inclusion criteria were any cat with a complete case record (owner data, patient signalment and anamnesis, complete clinical findings and cardiac investigation) and an echocardiographic diagnosis of RCM.
The diagnosis was based in both institutions on the echocardiographic presence of left atrial (LA)/biatrial enlargement, normal LV wall thickness (m-mode LV wall thickness in diastole <6 mm measured by the leading edge-to-leading edge method), normal or mildly decreased systolic function and restrictive LV filling pattern with pulsed wave Doppler echocardiography (E wave/A wave ratio [E/A] >2).2,4,8 The latter criteria was not strictly considered for inclusion in cases where E and A waves were summated (for tachycardia) or A wave was absent (owing to supraventricular arrhythmia) and all previously mentioned echocardiographic characteristics were present. When focal hypertrophy was detected (>6 mm in M-mode or B-mode measurements) the case was excluded from the study.
LA enlargement was defined by a LA to aortic root (LA/Ao) ratio greater than 1.5 on B-mode. 11 LA enlargement was subsequently classified as mild to moderate if the LA/Ao ratio was 1.5–2.0, whereas cats with a LA/Ao ratio >2.0 were considered to have severe LA enlargement. 11 Echocardiographic signs of increased risk for arterial thromboembolism (ATE) included the presence of spontaneous echocardiographic contrast (‘smoke effect’) or the direct visualisation of intracardiac thrombi in the LA or auricle.
Cats diagnosed with a CM other than RCM, congenital heart disease, systemic hypertension, hyperthyroidism or those with incomplete case records were excluded from the analysis.
Systemic systolic blood pressure (BP) was assessed non-invasively using a Doppler-based technique in all cats, as recommended by the American College of Veterinary Internal Medicine Guidelines. 12 When BP was >150 mmHg on serial repeated measurements, the cat was classified as affected by systemic hypertension and excluded from the study. 13 All cats older than 10 years of age had their thyroxine (T4) levels tested. 14 If the cat presented with a clinical history or with clinical findings related to the presence of hyperthyroidism (polyphagia, progressive weight loss), T4 levels, haematology and biochemistry were performed regardless of the cat’s age. Thoracic radiographs were performed in all cats with respiratory distress.
Respiratory distress was defined by the presence of increased respiratory rate associated with an increase in effort and/or open-mouth breathing and/or orthopnoea. Increased respiratory rate (tachypnoea) alone was not considered sufficient owing to the possibility of tachypnoea being identified in normal cats in the hospital environment. 15
Follow-up status and cause of death was determined by reviewing the medical records and/or telephone interviews with the owners by investigators or trained senior veterinary students, when more information was required. If the cats had died, an attempt was made to classify the events as cardiac related or not. Cardiac-related death was defined as death occurring because of progression of clinical signs of heart failure (HF)/ATE. Euthanasia because of refractory HF/ATE was scored as cardiac-related death. Sudden death was regarded as cardiac-related if no other cause of death was obvious. Cats still alive or that had died or were euthanased for reasons unrelated to cardiac disease were censored in the statistical analysis. Cats lost to follow-up were included in the survival analysis up until the last time point at which they were known to be alive and were thereafter censored in the analysis.
Statistical analysis
Basic descriptive statistical analyses were performed using Microsoft Excel. Data were analysed using a commercially available software (SPSS Statistics for Windows v23). In all cases a P value <0.05 was considered significant. The Shapiro–Wilk test was used to verify normal distribution of variables. Normally distributed data are reported as mean ± SD and non-normally distributed data as median and interquartile range (IQR).
Survival time was calculated from the time of diagnosis to the date of death or last telephone contact. The Kaplan–Meier method was used to estimate the survival function and plot time to event curves in the different group. A log-rank test with right censoring was used to determine whether a significant difference existed between groups.
Schoenfeld residuals and time-dependent covariates were used to test the assumption of proportional hazards. Univariate and multivariate Cox proportional hazard analysis were performed in order to determine the effect of any variable on survival. Hazard ratio (HR) and 95% confidence intervals (CIs) were calculated.
Variables were added to the multivariable model in a manual stepwise manner, first including all variables statistically significant in the univariate analysis and then excluding those not reaching statistical significance one by one until all the variables included were statistically significant (backwards regression analysis).
Variables assessed for their effect on outcome were breed (longhair vs shorthair), sex, age at presentation, presence of clinical signs (respiratory distress, syncopal episode, limbs paresis/paralysis) presence of pleural/pericardial effusion, pulmonary oedema and abdominal distension, and echocardiographic variables (LA/Ao ratio, mild-to-moderate or severe LA enlargement, LV fractional shortening (FS), LV dimensions in systole/diastole and presence of echocardiographic signs of increased risk for ATE).
Results
From January 1997 to December 2015, 767 cats were diagnosed with a CM; most cats had HCM (n = 594; 77.4%), 115 cats (15.0%) were classified as having RCM, 34 cats (4.4%) as having dilated CM (DCM), 22 cats (2.9%) as having unclassified CM (UCM) and four cats (0.5%) as having arrhythmogenic right ventricular CM (ARVC). Twenty-five cats with RCM were thereafter excluded because they did not meet the inclusion criteria (12 incomplete case records and 13 incomplete echocardiographic reports). The final study population comprised 90 cats with an echocardiographic diagnosis of RCM.
Male cats were predominant in the population (58.9% male, 41.1% female). Forty-three males and 34 females were neutered. The breed population included mostly domestic shorthair cats (n = 60; 67%), followed by Persians (n = 15; 17%), longhair cats (n = 11; 12% [four Norwegian Forest Cats, four Birmans and three Maine Coons]), three Siamese cats and one Chartreux. At presentation, mean age was 10.0 ± 4.3 years and median weight was 3.8 kg (IQR 3.2–5 kg). The majority (n = 87; 97%) of cats had clinical signs at presentation, with only three cats being asymptomatic. Only the minority of cats had a murmur (n = 9; 10%). Twelve cats (13%) presented supraventricular arrhythmias.
Presenting complaints are listed in Table 1.
Presenting complaints in 90 cats with RCM. Note that some cats showed more than one presenting complaint
Seventy-five cats were presented because of respiratory distress, six of which also had limb paresis/paralysis and eight also abdominal distension
Thoracic radiographs were performed in the 75 cats (83%) presenting with respiratory distress. Pleural effusion was observed in 44 cases (59%), pulmonary oedema in 19 (25%) and both in 12 cases (16%).
All cats received treatment with furosemide and angiotensin-converting enzyme (ACE) inhibitors. Diltiazem was administered in all cats that presented with supraventricular arrhythmias. Anti-thrombotic treatment (low-dose aspirin before January 2013 or clopidogrel thereafter) was administered in all cats with clinical and/or echocardiographic signs of increased risk of ATE (presence of thrombus, smoke effect) and/or moderate atrial dilatation, (LA/Ao >1.8).
All cats included in the study were conscious, unsedated and manually restrained during the echocardiographic examination. Echocardiographic parameters in cats with RCM are shown in Table 2. All cases had a restrictive pattern with the exception of 18 cats: six cats had the E wave summated to the A wave for tachycardia, and 12 cats had supraventricular arrhythmia and the A wave was absent. Most cats had severe LA enlargement (n = 72; 80%) and 25 cats (28%) presented echocardiographic signs for increased risk of ATE (smoke effect or mural thrombi). Patchy or extensive areas of increased echogenicity of the endocardium were observed in only two cases. During the echocardiographic examination pericardial effusion was observed in six cats.
Echocardiographic parameters in 90 cats with RCM
LVFWd = left ventricular free wall in diastole; IVSd = interventricular septum in diastole; LVIDd = left ventricular internal dimension in diastole; LVIDs = left ventricular internal dimension in systole; FS = fractional shortening; LA = left atrium; Ao = aorta; E wave = transmitral early (E) filling velocity; A wave = transmitral atrial (A) filling velocity
Follow-up was available for 60 cats and the median survival time (MST) was 69 days (95% CI 0–175). Cardiac-related death occurred in 50 cats (83%), five cats were still alive at last follow-up and five cats had died from unrelated cardiac causes (three from neoplasia and two from chronic kidney disease).
In the univariate Cox analysis, respiratory distress, pleural effusion and LA enlargement (mild-to-moderate vs severe) showed statistically significant effects on survival with HRs, respectively, of 3.54 (95% CI 1.25–9.99; P = 0.017), 2.34 (95% CI 1.16–4.71; P = 0.017) and 2.32 (95% CI 1.08–4.99; P = 0.031). In the multivariate Cox regression backward analysis, only respiratory distress showed a statistically significant effect on survival. Cats presenting without respiratory distress showed a MST of 466 days (95% CI 0–1208), in contrast with cats that presented with respiratory distress showing a shorter MST of 64 days (95% CI 8–120; P = 0.011) (Figure 1).
Kaplan–Meier survival curves: cats presenting without respiratory distress showed a median survival time (MST) of 466 days (95% confidence interval [CI] 0–1208) vs cats that presented with respiratory distress showing a shorter MST of 64 days (95% CI 8–120; P = 0.011)
Discussion
The results show that RCM is almost exclusively diagnosed at a late stage when cats are referred after the development of clinical signs. Additionally, long-term prognosis is poor. Almost all cats in our study showed signs of congestive HF alongside cardiogenic thromboembolism, similar to what has previously been reported.2,3,16
Prognosis is poor, as shown by the short survival time after the diagnosis: the MST in our study population was slightly shorter (2 months) than the MST reported by Fox et al (3 months) 2 and Ferasin (4 months), 3 and longer than that reported by Kimura et al (1 month). 16 Survival seemed better in those cats not presenting with respiratory distress; however, they were a minority.
In our study respiratory distress was the only factor affecting survival in the multivariate analysis (stronger than pleural effusion and atrial enlargement); this emphasises how in a population of cats with RCM the presence of respiratory distress is the most useful variable in order to distinguish cats with poor prognosis.
In our study population, RCM was the second most commonly diagnosed CM in cats, with a 10 year prevalence of 15% in all cats with CM referred to our two centres. DCM and UCM were less common, with prevalences of 4% and 3%, respectively. At our institutions, diagnosis of DCM is based on the echocardiographic observation of LV end-systolic diameter >14 mm and a fractional shortening of <28% in M-mode; diagnosis of UCM is made by exclusion, on the basis of evidence of myocardial abnormality that does not fit into any of the recognised disease classifications. 3 A variable prevalence in RCM has been reported in cats, with Schober reporting the lowest prevalence of RCM 17 – 2.4% of all cases of primary feline CM (n = 450) diagnosed between 2007 and 2015 (91.5% HCM, 2.4% DCM, 2.7% UCM, 1.1% were ARVC). In a study performed by Fox et al, 2 the prevalence of cats with RCM was also low (5%). In another report from Japan the prevalence of RCM (endomyocardial form) was slightly higher, with a prevalence of 13% (327 autopsies from cats with heart disease in a 10 year period). 16 In contrast, in a retrospective study from the UK, which included 105 cats with primary CM, the prevalence of RCM was 21%. 3 These differences regarding the prevalence of RCM are probably related to different diagnostic criteria used. Both inter- and intra-observer agreement for myocardial disease classification in cats have also been reported to be poor. 18 Most authors would agree that there is marked overlap between CM categories and there is a possibility, although infrequently observed, of a change in CM (eg, an end-stage HCM with LV wall thinning may result phenotypically in more RCM/DCM than HCM); there are examples of families of cats (Norwegian Forests) that include individuals with HCM and RCM, or a mixed HCM/RCM phenotype.8,9 Moreover, restrictive ventricular physiology is not RCM-specific but rather occurs over a wide range of myocardial pathologies, end-stage in particular. In general, definitive qualitative and quantitative echocardiographic criteria and diagnostic cut-offs in the diagnosis of feline RCM are poorly defined and remain controversial, with only one study reporting objective data.2,17 Challenge is the diagnosis not only by echocardiography, but also by pathology because accepted objective histopathology criteria in the diagnosis of feline RCM are lacking. 17 Therefore, the prevalence may have been influenced by the different diagnostic criteria and classification in different studies. Geographical distribution may also be another source of prevalence variation between the studies.
None of the cats in this study showed transition into a different CM.
Several previous studies of RCM have shown a female predisposition, or equal predisposition.2,3,19 Similar to Kimura et al, 16 in our study cats with RCM were predominantly male. The most represented breed was domestic shorthair, which is the most common cat breed in Italy. 19
The mean age at presentation reflects the adult onset of the disease, as reported in the literature.2,3,16 The wide age range at diagnosis may reflect a wide disease onset, as is the case for HCM, or it could indicate the presence of different underlying pathogenesis leading to a common final echocardiographic appearance, as could be the case for acute myocardial damage, myocarditis, end-stage HCM or neoplasia. 8
In the present study, a heart murmur was rarely identified in cats with RCM, which is similar to what has been reported previously; the lack of a heart murmur is in line with current observations that the presence or absence of a heart murmur is not a useful screening tool in cats. 20
Endomyocardial fibrosis was identified only in two cats. It is possible that this type of RCM might have been under-diagnosed during echocardiographic examination owing to the location and size, the quality of the ultrasonographic equipment, the skill of the operator and the lack of cooperation of some cats.
Limitations of this study are mainly related to its retrospective nature. Diagnosis was based only on echocardiography and post-mortem examinations were available only in a few cases; no cardiac biomarkers were available. B-mode measurements of LV wall thickness were not available in all cases, but those were a minority. We chose a value of ⩾6 mm for the definition of LV hypertrophy based on previously published studies, and no allometric scaling was used in order to control the effect of the body weight.2,4,21 No other diastolic information was available in cats with summated E and A waves for tachycardia or absent A wave for arrhythmia. The distribution of CMs might have been biased by the echocardiographic criteria used in both referral centres where the study was carried out; however, no consensus on CM classification is currently available in veterinary cardiology for uniform classification of feline CMs. 18 No cat included in the study was previously diagnosed with HCM based on a previous echocardiograph; we excluded cases in which focal hypertrophy was present in an attempt to exclude end-stage HCM. Nevertheless, we cannot completely rule out the possibility that some cats with HCM could have been included. Treatment in the current population was not standardised but consisted mainly of loop diuretics, ACE inhibitors and anti-thrombotic treatment. Finally, owner-related information could have biased the results owing to misinterpretation of clinical signs or failure to recognise cardiac-related death.
Conclusions
RCM can be considered an end-stage condition associated with a poor prognosis, with few cats not showing clinical signs and surviving longer than 1 year. Most cats died as a result of cardiac disease within a very short time.
Footnotes
Conflict of interest
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.