Radiographic and anatomical morphometric assessments of heart size in presumed healthy pet guinea pigs

Abstract Cardiac disease in guinea pigs has been reported in the literature; however, reference intervals for normal radiographic heart size obtained using objective measurement methods have not been provided for this species. The aim of this prospective, reference interval study was to describe cardiac dimensions in presumed healthy guinea pigs using the vertebral heart scale (VHS) from thoracic radiographs, as described for dogs and cats. Furthermore, an anatomical study was carried out to compare the radiographic and anatomical findings. Thoracic radiographs were acquired in right lateral recumbency for 30, client‐owned, conscious, presumed healthy guinea pigs and radiographs were acquired in left lateral recumbency for 10 presumed healthy guinea pigs as comparisons. The influence of sex, age, body weight (BW), and recumbency on the VHS and absolute cardiac measurements was investigated. The median (interquartile range; IQR) VHS was 7.4 (7.1‐7.6). No differences emerged between the VHS measured in right versus left lateral recumbency (P = .41) or between sexes (P = .16). The VHS values were not influenced by age (P = .53) or BW (P = .26). The anatomical study was carried out on 10 guinea pig cadavers, and in situ and ex situ cardiac measurements were taken using a caliper. A median (IQR) 7.5 (7.2‐8.0) VHS was assessed by this anatomical study. The reference intervals provided should be useful tools in the future for the radiographic interpretation of cardiac size in guinea pigs in clinical practice.


INTRODUCTION
Cases ofpet guinea pigs (Cavia porcellus) presented having signs indicative of cardiac disease in clinical practice have been reported in the literature. [1][2][3][4][5] Common clinical signs of cardiac disease are non-specific, such as dyspnea, lethargy, and anorexia. 6 It has been shown that only a minor percentage of guinea pigs with cardiac disease show heart murmurs on auscultation and that, on echocardiography, dilated cardiomyopathy is most frequently diagnosed in association with secondary pericardial effusion. 5 Instead, hypertrophic cardiomyopathy and valvular disease, in association with pleural effusion, are evidenced in a minority of cases. 5 Cardiomegaly, pleural effusion, and pulmonary edema are often present on thoracic radiographs in cardiopathic patients. 6 Standard thoracic radiography is the diagnostic method of choice in small exotic mammals for the assessment of the trachea, the lung parenchyma, and the pulmonary vessels. However, it is also routinely utilized for a first evaluation of the heart and the large blood vessels as well as for monitoring heart size and cardiac chamber changes over time and/or in response to treatment, 7 together with the utilization of more advanced and specific diagnostic techniques, such as ultrasonography, which should be performed in a complete cardiological assessment. 8 The guinea pig thorax is short, and the precordial section is only one-to-three intercostal spaces wide. 8 The heart occupies a relatively large space in the thorax, and it lies along the midline from the second to the fourth intercostal space. 8,9 The vertebral heart scale (VHS) system, as described for dogs by Buchanan and Bucheler,7 is an objective standardized diagnostic technique that is well established for dogs of various breeds 7,10-12 and cats, 13 but is being increasingly utilized in exotic companion mammals as well. To date, it has been used for the cardiac assessments of rabbits, 14,15 chinchillas, 16 ferrets, 17 mice, 18 and African hedgehogs, 19 among others. The accuracy of thoracic radiography in detecting deviations in the cardiac silhouette relies on the availability of speciesspecific normal reference data. In guinea pigs, however, the reference intervals for normal heart size derived from objective radiographic measurement methods have not yet been provided, causing clinicians to subjectively assess the presence of abnormal changes in the cardiac silhouette in clinical practice; for instance, associated signs, such as tracheal elevation, are currently being used as an aid for determining the presence of cardiac enlargement from thoracic radiographs. 6 The availability of species-specific reference ranges of normality regarding heart size could allow helping with the objective radiographic interpretation of the cardiac silhouette.  Maximal diameter of the caudal vena cava (CVC) and Length of the vertebra dorsal to the tracheal bifurcation (T5).
The measurement technique from the right lateral view is illustrated in Figure 1.

Anatomical study
An anatomical investigation of the guinea pig heart was carried out for comparison with the radiographic findings acquired in the present study and with the echocardiographic measurements reported in a previously published study. 20

STATISTICS
The statistical analyses were carried out by a veterinarian with PhD-

Radiographic study
The conscious radiographic procedure was well tolerated by all the guinea pigs examined. No complications, such as respiratory distress, or injuries associated with physical restraint were observed.
On the right lateral recumbency radiographs, the heart extended from the caudal border of the second rib to the fourth intercostal space   ( Figure 4B). Table 1 shows the statistics relating to all the signalment data and radiographic cardiac measurements taken from right lateral radiographs.

Anatomical study
All hearts appeared macroscopically normal. The hearts of those guinea pigs positioned in right lateral recumbency extended from the second rib to the fifth intercostal space (n = 1), the fifth rib (n = 3), or the fourth intercostal space (n = 6). In left lateral recumbency, the heart had a more cranial topography, extending from the second to the fourth ribs (n = 7) or the fourth intercostal space (n = 3).

F I G U R E 3 Histogram depicting the distribution of the VHS values measured on thoracic radiographs of 30 clinically normal pet guinea pigs
The VHS score obtained by anatomical dissection was not significantly different than that obtained by radiography (P = .3). The anatomically derived measurements obtained from 10 guinea pig hearts, along with the corresponding echocardiographic measurements obtained from a previous study carried out on guinea pigs, 20 and the P-values resulting from statistical comparison, are reported in Table 2.

DISCUSSION
The present study provides the first published report describing normal cardiac size assessment from thoracic radiographs using the VHS, together with an anatomical comparison, in presumed healthy pet guinea pigs. Species-specific radiographic reference intervals of normality could prove useful in clinical practice in the objective evaluation of the cardiac silhouette of guinea pigs, which typically tend to present unspecific symptoms at the onset of the pathology. 5 Radiographically, the majority of the hearts extended either to the fifth rib (30%) or to the fifth intercostal space (33.3%), therefore going beyond the caudal limit previously described in the literature (4th IS). 9 Moreover, radiographically, only 20% (6/30) of the hearts occupied two ISs, as previously reported 9 ; on the other hand, the majority (24/30; 80%) of the cardiac silhouettes occupied 2.5 to 3.5 ISs. The cranio-caudal limits and number of ISs occupied by the guinea pig heart herein reported could prove useful for at-a-glance heart size assessment before obtaining more objective cardiac measurements using the VHS. 8 No differences were observed between measurements taken in right versus left lateral recumbency. This finding is in line with studies carried out in chinchillas, 16 agoutis, 26 ferrets, 17 laboratory mice, 18 and dogs. 7 However, it is in contrast with studies carried out on some dog breeds including beagles 11,12 in which the VHS was found to be significantly larger in right lateral recumbency, and with studies carried out on rabbits 15 and whippets 10 in which measurements were significantly larger in the left lateral view.
The results of the radiographic VHS obtained from right lateral projections in the guinea pigs in the present study (7.30 ± 0.4) were lower than those reported for other rodents, such as mice (9.1 ± 0.5), 18 blackrumped agoutis (8.0 ± 0.31; 7.68 ± 0.41), 26,27 and chinchillas (8.90 ± 0.72), 16 but were also lower than the values reported for non-rodents, such as African hedgehogs (8.16 ± 0.48), 19 and dogs (9.7 ± 0.5). 7 It, however, is fairly consistent with the ranges reported for rabbits (7.55 ± 0.38; 7.60 ± 0.39), 14,15 cats (7.5 ± 0.3), 13 and black-tailed prairie dogs (7.12 ± 0.42), 28  It is important to note that normal VHS values do not necessarily rule out heart disease, as is the case with concentric hypertrophic dilatation, for instance. 7 In chinchillas, 16 ferrets, 29 and cats, 30 the VHS method proved to be scarcely sensitive, and only moderately accurate, in detecting selected cardiac abnormalities, such as cardiomegaly, or, in dogs, specific cardiac chamber enlargement. 31 The CVC maximum diameter tended to be lower than the length of the vertebra dorsal to the tracheal bifurcation (T5) in the majority of the guinea pigs, confirming what had been observed in dogs. 7 Moreover, the CVC/T5 ratio calculated in the present study was consistent with that reported for dogs. 7 Therefore, it could be hypothesized that, as is the case with dogs, T5 could be taken as the upper normal limit for the CVC in healthy guinea pigs. The CVC/T5 ratio is a useful parameter in the assessment of the intravascular volume status of the patient; an increased ratio has been seen to be suggestive of right-sided congestive heart failure in dogs. 32 Interobserver variability was assessed to be moderate-to-good for all the radiographic indices in the present study. Potential sources of variation included the individual observers' selection of cardiac and vertebral anatomic landmarks, and the transformation of the measurements of the cardiac long and short axes into units of vertebral length. 33 From the anatomical study, it emerged that the cardiac cranial and caudal extension limits were, in the majority of cases (60%), consistent with those reported in the literature, that is from the second to the fourth IS, whereas, in four cases (40%), the heart had a more caudal LA, apicobasilar length of the heart; SA, width of the heart at its widest point perpendicular to LA; T4, fourth thoracic vertebra extension than that previously described. 9 The resulting median VHS score, assessed by anatomical dissection, was not statistically different from the corresponding median radiographic value, which pointed out the similarity between the anatomical and the radiographic techniques. On the other hand, the anatomical measurements of the aortic and left atrial internal diameters were significantly less than those of the echocardiographic measurements at end-systole and end-diastole reported in the literature. 20 A possible explanation for the discrepancy was the post-mortem change in the myocardium, leading to modifications of the chamber diameters, 34 together with the loss of pressure and the consequent collapse of vessels such as the aorta. 35 Another explanation might be the lower mean BW of the guinea pigs dissected.
In fact, the above-mentioned echocardiographic study showed a positive correlation between BW and the left atrial diameter. 20 In addition, the LAD/Ao ratio calculated was almost double that assessed by echocardiography. 20 This finding was probably ascribable to false left atrial dilatation secondary to postmortem myocardium deterioration.
On the other hand, the internal diameters of the left and right ven- A limitation of the present study was the small number of guinea pigs enrolled; a larger number (>120) would be necessary to establish normal reference intervals. 23 Moreover, the radiographs in left lateral recumbency were obtained from ten guinea pigs only. The choice of using only manual restraint to obtain the radiographs was due to the fact that some anesthetic agents are known to alter cardiac measurements. 36 The implementation of X-ray positioning devices to decrease restraint stress, as well as the rapidity of the procedure and the absence of handling-related complications, were at the base of such a decision. It is, however, advisable that guinea pigs with possible cardiorespiratory disease be sedated for radiography. Another limitation of the present study was the lack of additional diagnostic investigations, such as echocardiography, to ascertain the actual lack of cardiac disease in the guinea pigs studied as it has been demonstrated that guinea pigs with cardiac disease tend to be present few or unspecific symptoms. 5,6 Authors, indeed, cannot completely exclude the possibility that some of the clinically healthy animals could have had subclinical disease. However, based on the absence of clinical symptoms on physical examination and cardiac auscultation performed by a veterinarian with expertise in exotic small mammal medicine, as well as the absence of radiographic signs of cardiopulmonary disease in the totality of the animals included in the study, the guinea pigs were assumed by the authors to be likely free of cardiac disease.
Therefore, also considering the radiographic differences in the cardiac silhouette seen between the presumed healthy guinea pigs and the two animals diagnosed with cardiac disease, the authors believe that the findings from this sample of clinically healthy guinea pigs can be valid for use as normal reference interval measurements. Additional studies involving healthy guinea pigs versus guinea pigs diagnosed with cardiopathy, are needed to strengthen the present findings and assess the accuracy of the VHS in detecting cardiomegaly in this species.
In conclusion, this study showed the feasibility of radiographic cardiac assessment in conscious guinea pigs and provided normal radiographic reference intervals which could prove useful for clinical use in the objective evaluation of cardiac size in guinea pigs.