Is cardiac troponin I accurate as a screening test for the diagnosis of HCM in cats | VETgirl Veterinary Continuing Education Podcasts
In this VETgirl online veterinary CE podcast, we review whether cardiac troponin I is an accurate screening test for the diagnosis of Hypertrophic Cardiomyopathy (HCM) in cats. Hypertrophic cardiomyopathy is defined as concentric hypertrophy of the left ventricle in the absence of any other identifiable cause, in particular systemic hypertension and hyperthyroidism in cats. Despite the relatively common occurrence of congestive heart failure (CHF) and arterial thromboembolism (ATE) in cats with HCM, definitive diagnosis prior to these outcomes poses a diagnostic challenge. Echocardiogram remains the gold standard for diagnosis, but may not be feasible in many cases due to need for specialized equipment and training, or cost. Biomarkers such as N-terminal pro-B type natriuretic peptide (NT-proBNP) and cardiac troponin I (cTnI) have undergone extensive investigation in cats, with a correlation between elevated values of these markers and HCM now established. The predictive value of cTnI as a global screening test for HCM remains unclear, however.
So, Hertzsch et al out of Munich, Germany, wanted to evaluate this in a study entitled Evaluation of a sensitive cardiac troponin I assay as a screening test for the diagnosis of hypertrophic cardiomyopathy in cats. The authors wanted to determine the utility of, and appropriate cutoff value for, a sensitive cTnI assay in screening cats for HCM in a prospective study. Factors affecting cTnI concentrations and correlations between cTnI and echocardiographic variables were also investigated. Exclusion criteria for the study included any identifiable disease that could cause concentric left ventricular hypertrophy, other forms of concurrent heart disease, and renal disease. Guidelines for echocardiographic diagnosis and quantification of HCM were in accordance with previously reported data. All cats received an examination, echocardiogram, and blood pressure. Blood pressure was considered normal if systolic pressure was < 160 mm Hg. Blood testing included serum urea + creatinine (all cats), and thyroxin (T4) (cats with hypertrophy), with T4 levels > 4.5 ug/dL exclusionary. Serum was frozen and stored for eventual cTnI evaluations once the most sensitive assay was available. The lowest detection limit of the cTnI assay was 0.006 ng/mL.
A total of 166 cats were included in the study (97 male, 69 female). European Shorthair represented the largest breed group. 87 cats were considered healthy, 60 had HCM, 15 cats were considered equivocal for HCM, and four cats had ATE. Healthy cats were significantly younger (median 2.75 years) than cats with mild, moderate, and severe HCM (median age 9.92 years, 11.46 years, and 6.38 years, respectively). Male cats were overrepresented in all groups except healthy cats and ATE cats. 24 cats had cTnI below the assay detection limit, of which 21 were in the healthy group. These 24 cats had cTnI levels of 0.006 ng/mL included in the statistical analysis as a default.
So, what’d the authors find in this study? There were no significant effects of body weight, sex, or age on serum cTnI concentration (although it is previously accepted that cTnI levels do generally increase slightly with age). Significantly lower median cTnI levels did occur for healthy (0.013 ng/mL) and equivocal (0.022 ng/mL) groups compared with all other cats. A significant difference was also noted in cTnI levels for the mild HCM group (0.1 ng/mL) vs the severe HCM group (0.760 ng/mL), with the latter group displaying higher levels. ATE cats, despite a small group number overall (4), had significantly higher median cTnI levels (6.413 ng/mL) than cats with mild or moderate (0.174 ng/mL) HCM. For the purpose of statistical analysis, the severe HCM group was subdivided into three additional groups: compensated (15 cats), decompensated (11), and compensated under treatment (8). The compensated group displayed significantly lower (0.31 ng/mL) median cTnI levels than the decompensated (4.11 ng/mL) and compensated under treatment (0.64 ng/mL) groups.
From an echocardiographic standpoint, cTnI was significantly, positively correlated with left ventricular free wall thickness in diastole. Positive correlations with interventricular septal thickness during diastole and left atrial-to-aortic root ratio (LA:Ao, a parameter for assessing left atrial size) were also observed. Negative correlations were identified between cTnI and end-diastolic and end-systolic left ventricular chamber dimensions.
Statistical analysis to determine cutoff values for cTnI was initially performed without inclusion of the equivocal group, but was also subsequently performed in two additional manners: with equivocal cats included in the healthy group, and with equivocal cats considered to be in the HCM group. This was done to strengthen the analysis by confirming that the exclusion of the equivocal group from the initial analysis did not have a significant impact on the overall results (which it did not). A cTnI cutoff of 0.06 ng/mL differentiated healthy cats from HCM cats with high sensitivity and specificity (91.7% and 95.4%, respectively). Positive predictive value for differentiating healthy cats from HCM cats was 77.4% at this cutoff value; negative predictive value was 98.5%.
The results of this study, when compared with prior studies of cTnI in cats with HCM, are important because they demonstrate that cTnI can in fact be a useful screening biomarker for detecting HCM in cats, even those with mild disease. Prior studies used first-generation cTnI assays, not a highly sensitive cTnI assay like that used in the current study, and those prior results suggested that cTnI may only be useful in detecting cats with moderate-severe HCM. The current study suggests that cTnI has utility in detection of cats with mild HCM. The present study also suggests that cTnI has utility in detecting cats with asymptomatic, but severe, HCM as compared with those with decompensated HCM, or HCM compensated only by virtue of cardiac therapy. cTnI in the present study was not particularly effective at differentiating moderate HCM from mild or severe forms (thus retains some limits in its ability to quantify severity). Markedly elevated cTnI levels were not specific, as they occurred with both severe HCM cats and cats with ATE.
The cTnI cutoff of 0.06 ng/mL appeared reliable at differentiating healthy cats from those with HCM with a high degree of confidence, keeping in mind that cats with other diseases were excluded from the study (and that these diseases, e.g. hyperthyroidism, systemic hypertension, renal disease can increase circulating cTnI levels). It’s important to recognize the limits of the clinical application of this study’s results, in that they suggest cTnI levels have utility in screening cats who are otherwise healthy for HCM but does not provide any data on how to extrapolate this data to cats with other diseases (which is important, given the relative prevalence of the hyperthyroidism, hypertension, and renal disease in aging cats). In comparison with NT-proBNP, cTnI shares the confounding influence of additional diseases that may artifactually increase levels. It currently has the advantage of a notably lower cost than NT-proBNP assays in some countries. But direct comparison of the two biomarkers is warranted to determine if one holds a distinct advantage in screening for mild HCM in cats, or perhaps if a combination of the two proves even more effective. Limitations of the study acknowledged by the authors include lack of serial cTnI measurements or echocardiography [thus inability to determine if any of the cats with “HCM” in fact had transient myocardial wall thickening (TMT)], lack of histopathology, and lack of investigation of the effect of common feline co-morbidities on cTnI levels in these cats.
So, what do we take from this VETgirl podcast? This is a useful, well-constructed study with a relatively narrow focus that enables straightforward interpretation of the data, and its potential clinical utility. Given the great number of feline patients who display no cardiac clinical signs and have no identifiable cardiac abnormalities on physical examination (e.g. murmur, gallop, arrhythmia) but who do, in fact, have cardiomyopathy, simple, inexpensive screening tests are highly desirable as they will help identify disease earlier, and also potentially limit the unnecessary expense of echocardiography in many cats who do not have heart disease. This study further supports that cTnI has its place at the table in pre-echocardiographic screening for HCM, and suggests it may have greater ability to detect mild HCM than previously suspected. Whether it is superior to other biomarkers, such as NT-proBNP, remains to be seen but in reality it may prove that a combination of these biomarkers (vs. either one alone) is superior in screening for cardiomyopathy. I do highly recommend that cardiac biomarker testing (NT-proBNP, cTnI, or both) be included in routine screening blood panels [e.g. complete blood count, chemistry profile, and thyroxine (T4)] for middle and older-aged cats, given the paucity of cardiac-specific enzymes on these routine panels.
1. Hertzsch S, Roos A, Wess G. Evaluation of a sensitive cardiac troponin I assay as a screening test for the diagnosis of hypertrophic cardiomyopathy in cats. J Vet Intern Med 2019;33: 1242-1250.