Right Ventricular Epicardial Fibrosis in Mice With Sternal Segment Dislocation

Abstract

We report coincident sternal segment dislocation and focally extensive right ventricular epicardial fibrosis observed during routine histopathology evaluation of C57BL/6N mice as part of a high throughput phenotyping screen conducted between 4 and 16 weeks of age. This retrospective case series study was conducted to determine whether cardiac fibrosis was a pathological consequence of sternal segment dislocation. We identified sternal segment dislocation in 51 of the total 1103 mice (4.6%) analyzed at 16 weeks of age. Males were more frequently affected. In all cases but 2, the dislocation occurred at the fourth intersternebral joint. In 42 of the 51 cases (82.4%), the dislocation was encased by regenerative cartilaginous callus that protruded internally into the thoracic cavity (intrathoracic callus) and/or externally to the outer aspect of the sternum (extrathoracic callus). Displacement of dislocated ends of the sternum into the thoracic cavity was present in 19 of 51 cases (36.5%). Coincident minimal or mild right ventricular epicardial and subepicardial fibrosis was observed in 22 of the 51 cases (43%) but was not observed in any of the mice in the absence of sternal segment dislocation. Our data suggest that right ventricular fibrosis was likely caused by direct injury of the right ventricle by the dislocated ends of the sternum and/or by intrathoracic callus that develops post dislocation. Potential pathogenesis for the sternal and cardiac lesions and their implication for the interpretation of phenotypes in mouse models of cardiopulmonary and skeletal disease are discussed.

Introduction

In humans, traumatic injuries to the chest are associated with sternal fracture or sternal segment dislocation.^23,31 Sternal fracture involves disruption of the cortex of the sternum and most commonly results from motor vehicle impact and less commonly from falls and direct violence.⁵ Insufficiency sternal fractures secondary to osteoporosis have also been described in elderly women.^7,12 Serious complications of sternal fracture include various cardiothoracic injuries such as pulmonary contusion, myocardial contusion and laceration, and aortic injury.^8,20,33 The right ventricle is more frequently involved⁸ since the sternal boundary of the heart consists of the free surface of the right ventricle. In contrast to sternal fracture, sternal segment dislocation (SSD) involves separation of the sternebrae at the intersternebral cartilaginous joint/junction. Similar causes as sternal fractures are implicated in SSD.^18,30 It is an extremely rare pathology in humans and mostly reported in children.^23,31 The sternum is less susceptible to fracture in children since the sternum and associated structures are more elastic and the ribs are more flexible than in adults.³¹ Further, the sternebrae in early childhood are joined by primary cartilaginous joints (synchrondrosis) while they are fused by synostosis in adulthood.³¹ Unlike humans, the sternebrae in rats and mice do not fuse and remain separated by intersternebral cartilaginous joints/junctions. To our knowledge, sternal fracture or dislocation has not been reported in laboratory rodents. Reported sternal lesions have been limited to degenerative disease of the intersternebral cartilaginous joints in aging rats³³ and mice.^16,29

We have observed coincident presence of SSD and right ventricular epicardial and subepicardial fibrosis (RVEF) in a subset of mice that have undergone comprehensive phenotype screening. More specifically RVEF was exclusively observed in mice with SSD that were either associated with large reparative cartilaginous callus and/or displacement of dislocated ends of sternebrae into the thoracic cavity. In order to investigate the association between SSD and RVEF, we carried out a retrospective histopathological review and statistical analysis of all cases of SSD from mice that have been subject to a similar battery of phenotyping tests. Our findings strongly suggest that RVEF was likely caused by direct abrasion or laceration of the right ventricle by dislocated and displaced sternebrae and/or by large reparative cartilaginous callus that protruded into the thoracic cavity. The incidence of SSD should be a consideration when interpreting results in mouse models of skeletal and cardiovascular disease. Further, we have suggested potential mechanisms for SSD and proposed remedial measures to minimize the incidence of this lesion in experimental mice processed through phenotyping pipelines.

Methodology

Ethics Statement

The care and use of all mice in this study was carried out in accordance with UK Home Office regulations, UK Animals (Scientific Procedures) Act of 1986, and the Canadian Council on Animal Care (CCAC) guidelines for Use of Animals in Research and Laboratory Animal Care under protocols approved by the Toronto Centre for Phenogenomics Animal Care Committee (ACC).

Animals and Phenotyping Pipeline

A retrospective analysis of the mouse pathology database at the Toronto Centre for Phenogenomics was conducted on 1103 C57BL/6N mice (107 wild type and 996 mutants) submitted for histopathology phenotyping as part of a multitest phenotyping pipeline; typically 4 mice (2 females and 2 males) are analyzed by histopathology (Fig. 1). The 996 mutant mice belonged to a total of 234 mutant lines generated from targeted C57BL/6N embryonic stem cells.^4,25 The 107 wild type mice were co-raised age-matched controls generated from in-house C57BL/6N breeding colonies. Of the total 1103 mice analyzed, 900 (450 females and 450 males) were phenotyped by the Mouse Genetics Project (MGP) at the Wellcome Trust Sanger Institute (UK), and the remaining 203 mice (107 females and 96 males) were phenotyped by the Toronto Centre for Phenogenomics (TCP) (Canada). The dietary background of the mice used in the study is varied. Of the 900 mice from the Sanger MGP, 841 were on high fat diet challenge (21.4% fat by crude content; 42% calories from fat, 43% from carbohydrate, and 15% from protein) (Special Diet Services Western RD 829100, SDS, Witham, UK). The remaining 59 mice from the Sanger MGP were on mouse breeder diet (9% fat by crude content; 55% calories from carbohydrate, 21.5% from fat, and 23.5% from protein) (Mouse Breeder Diet 5021 from Lab-Diets, London, UK). All mice from TCP were on regular diet (6.2% fat by crude content; 58% from carbohydrate, 18% from fat, and 24% from protein) (Diet number 2918; Harlan Teklad, Mississauga, Ontario). All mice had undergone a standardized battery of clinical phenotyping tests designed to interrogate different biological systems including reproduction, development, infection and immunity, musculoskeletal system (including X-ray imaging), metabolism, and endocrinology³²; http://www.mousephenotype.org/impress. Mice were not subjected to any invasive experimental procedure that involved the thorax. The typical workflow from clinical phenotyping to pathology analysis is presented in Figure 1.

Figure 1.

Clinical phenotyping pipeline and terminal histopathology. Clinical (blue) and pathology (green) phenotyping pipeline showing tests performed between 4 and 16 weeks of age. Seven female and 7 male mutant mice are routinely processed through this pipeline for each allele screened. Intensive handling of mice occurs at week 9 (solid arrow head) for modified SHIRPA and additional but less intensive handling (arrows) occurs at week 13 (glucose tolerance), week 14 (auditory brainstem response), and week 15 (ophthalmoscope). SHIRPA: SmithKline Beecham, Harwell, Imperial College, Royal London Hospital, phenotype assessment. Modified SHIRPA is a combined assessment of behavioral and morphological/physical abnormalities. *The majority of mice from Sanger MGP did not have open field test. CBC, complete blood count.

Histopathology

Mice were sacrificed at the end of the phenotyping pipelines (16 weeks of age), and a complete necropsy and comprehensive tissue collection for histopathology was done. Fresh tissues were immersion fixed in 10% neutral buffered formalin (NBF), embedded in paraffin in multi-tissue blocks, sectioned at 4 μm, and stained with hematoxylin and eosin (HE). The standard set of tissues routinely collected for histopathology, embedding, orientation, and sectioning procedures were described previously.¹ The stained tissue sections included the knee joint and a midsagittal section of the sternum extending from the manubrium (the first sternebra) to the xiphisternum and the xiphoid cartilage. The heart was sectioned longitudinally with view of all chambers in cases obtained from the MGP pipeline whereas 2 midtransverse sections were available in cases obtained from the TCP. Prior to embedding, all osseous tissues were decal-cified for 96 hours in TBD-2 Decalcifier solution (Thermo Scientific, Toronto, Canada). Selected sections of the sternum and the heart were cut from archived blocks and stained with Masson's trichrome to evaluate the extent of fibrosis.

To illustrate the spatial relationship between the sternum and the right ventricle, the entire thoracic cage (thoracic inlet to the diaphragm) was collected intact from a wild type 16-week-old C57BL/6N mouse. The thoracic cage was immersion fixed in NBF for 48 hours after small slits were made in the diaphragm to allow the fixative into the thoracic cavity. This was followed by decalcification for 96 hours as described previously. A midsagittal section of the whole thoracic cage was made by sectioning through the sternum and the thoracic vertebrae and photographed.

Histopathological evaluation focused on the sternum and the heart. Examination of the SSD included the presence or absence of displacement of dislocated segments and presence or absence of cartilaginous callus. Cases with cartilaginous callus were further classified based on the location of the callus (intrathoracic, extrathoracic, or both). The size of the intrathoracic callus was measured as the perpendicular length from the cortical surface of the dislocated sternebrae to the tip of the callus within the thoracic cavity. Based on this linear measurement, the size of the callus was further designated as large if the length of the callus was greater than the median value. Callus size was designated as small if the linear distance was smaller or equal to the median value (see the following for histopathology description for these classifications). The presence or absence of RVEF was documented for each case.

Statistical Analysis

A logistic regression model was used to explore the association between displacement of the dislocated sternal segments or the size of the intrathoracic callus and the occurrence of RVEF. The outcome was computed as odds ratio (OR), which estimates the ratio of occurrence to nonoccurrence of RVEF. Chi-square (χ²) analysis was used to compute the sex-based risk for SSD. All tests were 2-tailed with a significant level at P < .05. Data analysis was performed using the STATA Intercooled 11 statistical software (Stata Corporation LP, College Station, TX, USA).

Results

Sternal Lesions and Relevant Clinical Phenotypes

Nine of the 1103 mice (0.82%) reviewed in this study had abnormal sternum X-ray. An example of normal and abnormal sternal X-ray images is shown in Figures 2a and 2b, respectively. Gross abnormality of the sternum was not reported in any of the mice during necropsy. On microscopic examination, 51 of the 1103 mice (4.62%) analyzed had SSD. Of the 51 mice with SSD, 38 were from Sanger MGP (4.2% of Sanger MGP mice) whereas 13 were from TCP (6.4% of TCP mice). Nine of the 51 mice (17.6%) with SSD had an abnormal X-ray. Of these 51 mice with SSD, 7 originated from a total of 107 wild type animals used as phenotyping controls (6.45%). The remaining 44 mice belonged to a pool of 996 mutant mice (4.41%); these 44 mice were distributed among 33 mutant lines. Only 3 of these 33 mutant lines (10.7%) had a reported skeletal abnormality from phenotype assessment. These include 1 line with decreased length of long bones and increased trabecular bone thickness, a second line with decreased bone mineral content and truncated ribs, and the third line had kyphosis (abnormal spine curvature), decreased bone mineral density, decreased bone mineral content, and decreased bone length.

Figure 2.

a. Normal sternum; 14-week-old unaffected mouse. The sternum has normal contour and joints. The fourth intersternebral joint is indicated by arrow head. Note normal position of the xiphisternum (XS). Figure 2b. Sternal segment dislocation; 14-week-old affected mouse. Dislocation of the joint between the fourth and fifth intersternebral joint is evident (arrow head). The dislocated ends of the fourth and fifth sternebrae are displaced into the thoracic cavity while the xiphisternum (XS) protrudes ventrally. Histopathology of the sternum from the same mouse at 16 weeks of age is included in Figure 3b.

Distribution by sex showed that 16 of the 557 females (2.87%) and 35 of the 546 males (6.41%) had SSD. Males were more likely to have SSD (χ² = 7.83; P = .005). Table 1 shows the distribution of cases of SSD by sex and lesion characteristics including presence of coincident cardiac lesion.

Table 1.

Distribution of Cases of Sternal Segment Dislocation by Sex, Callus Location, Callus Size, Sternal Displacement, and Coincident Right Ventricular Epicardial and Subepicardial Fibrosis.

	Sternal Segment Dislocation	Right Ventricular Epicardial and Subepicardial Fibrosis
Total number of cases	51	22
Females	16	7
Males	35	15
No callus	9	2a
Callus	42	20
Intrathoracic and extrathoracic callus	38	19
Extrathoracic callus only	4	1a
Small intrathoracic callusb	26	10
Large intrathoracic callusc	12	9
Sternal displacement/misalignment	19	12

^aMouse also had sternal displacement/misalignment.

^bSmall intrathoracic callus ≤ median callus size (480 μm).

^cLarge intrathoracic callus > median callus size (480 μm).

In 49 of the 51 cases (96%), SSD occurred at the fourth intersternebral cartilaginous joint (the joint between the fourth and fifth sternebra). In 1 case, SSD occurred at the joint between the fifth sternebra and the xiphisternum (xiphisternal junction). In another case, the manubriosternal joint was involved (manubriosternal joint dislocation). The cortices of the affected sternebral segments were unremarkable, hence ruling out the diagnosis of sternal fracture. An example of normal sternum from an unaffected mouse (Fig. 3a) and sternal dislocation (Fig. 3b) are presented. In 42 of the 51 cases (82.4%), SSD was associated with nodular cartilaginous hyperplasia (cartilaginous callus) (Fig. 3b); these were considered old dislocations, and their histopathological features are described in the following. The callus originated from the periosteum of contiguous sternebrae and bridged the outer aspect of the dislocated segments (interpreted as periosteal bridging callus). Callus was not present in 9 of the 51 cases (15.7%); these were considered recent dislocations and their histological features are described in the following.

Figures 3. Sternal segment dislocation; 16-week-old mice; HE.

Figure 3a. Normal sternum from unaffected mouse. No notable abnormality was observed in the sternal contour and joints (arrow head). S4, fourth sternebra; S5, fifth sternebrae; PM, pectoral muscle; XS, xiphisternum; Xi, xiphoid. Figure 3b. Sternal segment dislocation at the fourth intersternebral joint (arrowhead) is enveloped by a periosteal bridging callus. Thoracic X-ray from the same mouse at 14 weeks of age is included in Figure 2b.

Two types of callus growth were observed. In the majority of the cases (38/42; 90.5%), the callus was present on both inner (intrathoracic) and outer (extrathoracic) sides of the affected joint (Fig. 4a). In the remaining cases (4/42; 9.5%), the callus was strictly extrathoracic protruding into the pectoralis musculature (Fig. 4b). The size of intrathoracic callus in the 38 cases ranged from 200 μm to 900 μm in length (median = 480 μm; 95% CI, 397–573). In 12 of the 38 cases (31.6%), the callus was large (bigger than the median value, 480 μm); these typically presented as prominent bulge or protrusions on the thoracic floor (Fig. 4a). In 26 of these 38 cases (68.4%), the callus was small (equal or smaller than the median value, 480 μm). In most of these cases, the callus presented as a slight bulge on the thoracic floor (Fig. 4c).

Figure 4.

a. Sternal segment dislocation with an enveloping intrathoracic and extrathoracic callus. The intrathoracic callus is large and protrudes up to 600 μM into the thoracic cavity. The extrathoracic callus (*) extends into the pectoral muscle (PM). An amphophilic amorphous material is present within the dislocated joint and extrudes dorsally into the lower aspect of the intrathoracic callus (arrow head). Figure 4b. Sternal segment dislocation with an extrathoracic callus. The extrathoracic callus (*) is segmentally ossified with marrow formation (arrow head). Intrathoracic callus is absent. Figure 4c. Sternal segment dislocation with an enveloping intrathoracic and extrathoracic callus. The intrathoracic callus is small and protrudes up to 300 μM into the thoracic cavity. The extrathoracic callus (*) is prominent and extends into the pectoral muscle (PM). The extrathoracic callus is beginning to ossify (arrow). An amphophilic amorphous material is present within the dislocated joint (arrow head).

In 19 of 51 cases (37.3%) with SSD, there was a displacement (misalignment) of the dislocated ends or edges with variable degree of protrusion into the thoracic cavity (Figs. 5a, 5b). Sternal displacement was also evident on X-ray in some of the cases (Fig. 5c). In most of the cases with sternal displacement, there is a significant inward angulation of the caudal fragment and overriding of the rostral fragment. In 16 of these 19 cases (84.2%) with sternal displacement, there was also an intrathoracic and/or extrathoracic callus; the remaining 3 cases did not have callus. These latter cases were accompanied by haemorrhage, minimal fibrin deposition, marked granulation, early fibroplasia, and a variable severity of neutrophilic infiltrate (consistent with recent dislocation) (Figs. 6a, 6b).

Figures 5. Sternal segment dislocation with displacement; 16-week-old affected mouse.

Figure 5a. Sternal segment dislocation at the fourth intersternebral joint (arrow) with displacement of the fourth and the fifth sternebrae into the thoracic cavity. S4, fourth sternebra; S5, fifth sternebrae; XS, xiphisternum; C, caudal; D, dorsal; R, rostral; V, ventral. HE. Figure 5b. Higher magnification of S4-S5 articulation in Figure 5a. Intrathoracic (dorsal) displacement of the fourth and fifth sternebrae (arrow) and an extrathoracic callus (*) are evident. Figure 5c. Thoracic X-ray from the same mouse in Figures 5a and Fig. 5b; dislocation at the fourth intersternebral joint with dorsal displacement of the sternebrae into the thoracic cavity (arrow) and a concomitant ventral protrusion of the xiphisternum (arrow head).

Figures 6. Recent sternal segment dislocation, caudal segment of the sternum; 16-week-old affected mouse; HE.

Figure 6a. Recent dislocation is evident by full thickness cleft along the intersternebral joint (arrow) and disruption of the sternal musculature and surrounding tissue. Note absence of callus formation. Figure 6b. Higher magnification of inset in Figure 6a; degeneration and necrosis of the sternal musculature, minimal hemorrhage, and inflammation are evident (arrow heads). The high magnification inset showing amphophilic granular material within the cleft (*) is taken from the area in the photo enclosed by the dashed lines.

All cases of SSD had amorphous amphophilic material along the joint line (Figs. 4a, 4c, 6a, 6b). There was minimal to moderate degeneration of the joint cartilage characterized by a decrease in metachromatic staining of the matrix (interpreted as loss of proteoglycan), fissures and cavities containing eosinophilic fibrillar material, and swelling and necrosis of chondrocytes. Clusters of proliferative chondrocytes (chondrones) were occasionally noted. The reaction in the surrounding connective tissue and associated sternal musculature ranged from haemorrhage, granulation tissue, and moderate numbers of neutrophils (recent dislocations) to extensive fibroplasia, fibrosis, and cartilaginous callus formation (older dislocations). In most cases, notably in recent dislocations, foci of regenerating multinucleated myocytes were present within the adjacent pectoral musculature. Moderate to large numbers of mononuclear inflammatory cells (mainly lymphocytes) were present in the older dislocations. In rare cases these chronic inflammatory cells expanded the overlying pleural lining.

A histopathological survey of the sternebrae in wild type C57BL/6N mice with no SSD showed that nearly 30% of the mice had early degenerative changes within the fourth intersternebral joint and to a lesser extent the third intersternebral joint (the joint between the third and fourth sternebrae). These changes were typically characterized by fissures and cavities within the joint cartilage, loss of ground substance, and occasional clusters of proliferating chondrocytes (Figs. 7a, 7b).

Figures 7. Early degenerative changes within the fourth intersternebral joint, caudal segment of the sternum; 16-week-old mouse; HE.

Figure 7a. Early degenerative lesion within the fourh intersternebral joint is evident by a fissure along the joint line (arrow heads). Figure 7b. Higher magnification of inset in Figure 7a; degenerative changes within the joint are characterized by cleft-like fractures and fissures (arrows), multifocal loss of matrix and chondrocytes resulting in cysts containing fibrous material (small arrow heads), and clusters of proliferating chondrocytes (large arrow head).

Right Ventricular Fibrosis and Association With Sternal Segment Dislocation

None of the mice had gross heart abnormalities. On microscopic examination, 22 of the 51 cases (43.1%) with SSD had RVEF. Five of the 13 SSD cases from TCP (38.5%) and 17 of the 38 SSD cases from Sanger MGP (44.7%) had RVEF. The RVEF was characterized by a 400 μm to 1500 μm wide and 50 μm to 250 μm deep plaque of fibrosis and/or fibroplasia on the epicardium and subepicardium of the right ventricular free wall (Figs. 8a, 8b) and occasionally at the apex of the right ventricle. Marked collagen deposition was evident within the superficial aspect of the lesion whereas immature/fine collagen fibers were seen within the deeper aspect of the lesion by Masson's-trichrome staining (Fig. 8c). Inflammation was invariably minimal and limited to small numbers of lymphocytes and macrophages and occasional neutrophils within the outer aspect of the fibrous plaque (Fig. 8c). The rest of the heart was unremarkable. The distribution of RVEF as a function of SSD and associated lesions is presented in Table 1. Of the total 22 cases with RVEF, 20 had callus whereas the remaining 2 did not have callus associated with SSD. Nine of the 12 cases (75%) with large intrathoracic callus and 10 of the 26 cases (38.5%) with small intrathoracic callus had RVEF. All of these 10 cases with small intrathoracic callus, and RVEF had concomitant displacement of the dislocated sternal segment. One case with RVEF had an extrathoracic callus only, but the dislocation was accompanied by a marked displacement of the affected segment into the thoracic cavity (Table 1). The likelihood of RVEF was significantly higher in mice with large intrathoracic callus compared to those with small intrathoracic callus (OR = 5.1, 95% CI, 1.11–23.37; P = .036). The likelihood of RVEF was even higher in cases with large intrathoracic callus when cases with coincident sternal displacement/misalignment were excluded (OR = 12.8, 95% CI, 1.7–97.2; P = .013). Twelve of the 19 cases (63.2%) with sternal displacement had RVEF. Three of these 19 cases (15.8%) with sternal displacement did not have callus (consistent with recent dislocations); 2 of these 3 cases had RVEF. Irrespective of presence or absence of callus, cases with sternal displacement/misalignment segments were more likely to have coincident RVEF than those without (OR = 3.6, 95% CI, 1.1–11.9; P = .036). In rare cases, mice with large intrathoracic callus had a focal depression on the right ventricular free wall accompanied by a corresponding mesothelial hypertrophy (not shown).

Figure 8. Epicardial and subepicardial fibrosis, right ventricle; 16-week-old mouse (the same case as in Fig. 4a).

Figure 8a. Epicardial and subepicardial fibrosis is present from the base to the middle aspect of the right ventricle (boundaries outlined by arrow heads). HE. Figure 8b. Higher magnification of inset in Figure 8a. A plaque of fibrosis and fibroplasia expands the epicardium and the subepicardium. Low numbers of mainly mononuclear cells are notable at the deeper aspects of the plaque of fibrosis (arrow heads). The underlying myocardium is unremarkable. HE. Figure 8c. Serial section from Figure 8b; collagen deposition is evident within the inner and middle aspect of the lesion (arrow heads). Masson's trichrome stain. LA, left atrium; LV, left ventricle; RA, right atrium; RV, right ventricle.

Spatial Relationship of the Right Ventricle and the Sternum

A midsagittal section of the entire thorax from a 16-week-old wild type mouse illustrates the spatial relationship between the sternum and the heart (Fig. 9). The sternal surface of the heart that is formed by the right ventricle is at its closest to the thoracic floor of the sternum between the third and fourth intersternebral joint.

Figure 9.

Thorax; midsagittal section; 16-week-old mouse. Close proximity between the ventral surface of the right ventricle and the sternal floor is indicated by asterisk. Formalin-fixed and decalcified tissue. LV, left ventricle; RV, right ventricle; S1-S5 (sternebrae 1–5); XS, xiphisternum.

Discussion

In this study, we have described a novel sternal lesion in mice and its association with right ventricular epicardial fibrosis. The pathological findings, physical proximity between the sternum and the right ventricle, and statistical analysis suggest that the RVEF was likely caused by injury of the epicardium by intrathoracic callus and/or an inwardly displaced sternal segment. RVEF was observed only in the presence of large intrathoracic callus or displacement of the dislocated sternal segments. Small callus was associated with RVEF only when there was also a coincident intrathoracic displacement of dislocated segments. Myocardial contusion and/or laceration occur in 6% to 12% of human cases with traumatic sternal fractures³³ and the right ventricle is more frequently involved.⁸ Callus formation following sternal fracture/dislocation has also been documented in humans.²³ A case of severe extrinsic pulmonic stenosis due to sternal callus was reported in 1 patient.²² This is not surprising as the sternocostal boundary of the heart in humans predominantly consists of the free surface of the right ventricle. This anatomical relationship is also maintained in the mouse as reported previously¹³ and as we have illustrated with a sagittal gross section of the entire thorax. This close proximity together with the delicate nature of the murine visceral pericardium¹⁵ make the right ventricle susceptible to injury by displaced sternebrae and/or by intrathoracic callus. Although unlikely, we also consider the possibility that the RVEF may be a direct extension of exuberant granulation and fibrosis originating from the sternal lesion.

This cardiac fibrosis was not observed in more than half of SSD cases. Most of these cases had small intrathoracic callus or minimal sternal displacement/misalignment. However, RVEF was also not evident in some cases despite the presence of large intrathoracic callus. It is possible that RVEF was missed in some cases because of the focality of the lesion. The orientation of heart section did not appear to be critical for the detection of RVEF because comparable frequency was observed by longitudinal section (Sanger MGP) and transverse section (TCP). The number of cases overall was not sufficient for statistical comparison between the two facilities.

Characterization of RVEF and Comparison With Cardiomyopathy

The focality and composition of the cardiac lesions were consistent with reactive (scarring) fibrosis of traumatic origin rather than primary degenerative change. Lesions suggestive of primary myocardial disease (cardiomyopathy) such as scattered areas of fibrosis, random multifocal myocardial degeneration and necrosis, or myofibre disarray were absent. Further, age-related cardiomyopathy is unlikely considering the young age of mice in our study (16 weeks old). A study in B6C3F1 control mice showed that cardiomyopathy was absent at 13 weeks of age and was found in only 10% of the mice at 2 years of age.¹⁴ The primary change in this age-associated cardiomyopathy in B6C3F1 mice was fibrosis of the left ventricular wall and interventricular septum; the right ventricle and atria were involved occasionally and only in severely affected mice.¹⁴ Fibrosis of the left ventricle also predominates in mouse models of hypertrophic cardiomyopathy.²⁸ In contrast, fibrosis of the right ventricular wall was reported as the predominant form of cardiac fibrosis in old C57BL/6J mice (22 to 86 weeks of age) in a series of studies from one laboratory.^17,26,27 These studies in senescent C57BL/6J mice showed patches of replacement fibrosis throughout the right ventricle free wall, but only in the sub-endocardium and mid-myocardium of the left ventricle.^17,26,27 This typical pattern of fibrosis in the right ventricle was associated with conduction slowing and arrhythmia.^17,27 The distinct pattern of replacement fibrosis within the right and left ventricles in these former studies is intriguing. Spontaneous cardiac lesion predominantly affecting the right ventricular free wall is well known in BALB/c and related C3H and DBA strains.^6,9 This lesion is typically characterized by dystrophic mineralization, a feature that was not observed in our case series. The cardiac lesion we described shares significant similarity to the “collagen plaques” that are commonly present over the anterior surface of the right ventricle in humans.²¹ These collagen plaques, also called milk spots or soldiers’ patches, represent focal epicardial fibrosis consisting of dense collagen occasionally accompanied by low numbers of underlying mononuclear cells.²¹ They are typically associated with right ventricular enlargement and they presumably arise from the impact of the beating heart with the subjacent surface of the sternum.²¹ Epicardial collagen plaques increase in incidence and size with age and with cardiac enlargement.²¹ In summary, the histopathological presentation of the cardiac lesion described in our report was consistent with scarring fibrosis.

Sternal Segment Dislocation: Potential Pathogenesis

Sternal lesions were almost exclusively located at the fourth intersternebral joint, suggesting an underlying predisposition of this joint for dislocation. Indeed, in this study, early degenerative changes such as cavitations were routinely observed in this joint in approximately 30% of the wild type C57BL/6N mice at 16 weeks of age (H. A. Adissu, unpublished observations) (Figs. 14, 15). Although not documented for the sternum, degenerative disease of other joints can occur as early as 2.5 months of age in some strains of mice.¹⁶ We speculate that degenerative changes may cause the fourth intersternebral joint to be synovial-like and mobile, hence the weakest and the most likely site to dislocate. In humans, the manubriosternal joint frequently appears synovial because of degeneration and cavitation of the fibrocartilage disc.²

We are not aware of underlying developmental or congenital sternum weakness or closure defects in the C57BL/6N mice from our study though it has been reported in a mutant line.¹⁹ In our study, underlying skeletal abnormality was seen only in 3 mice in the form of kyphosis, decreased bone mineral content, or bone mineral density. An underlying bone weakness in C57BL/6J inbred strain mice might have contributed to SSD. Bone mineral density and content in the C57BL/6J inbred substrain was shown to be the statistically lowest among 11 inbred strains.³ These bone phenotypes in the C57BL/6J inbred substrain likely hold true for the C57BL/6N inbred substrain we analyzed in this study based on recent comparison of these 2 substrains.²⁴ In humans, osteoporosis is a common predisposing condition to insufficiency sternal fractures and not surprisingly is seen more frequently in older women.^7,12 Furthermore, thoracic kyphotic deformity and subsequent compressive stress on the sternum is one of the causes of sternal fracture in humans.^7,12 Hence, physical stress, mild trauma, osteoporosis, or concomitant abnormality of the thoracic skeleton might contribute to SSD in the presence of degenerative disease of the sternebrae. We speculate that frequent animal handling and manipulation required in multi-assay phenotyping pipelines may play a role by inducing transient kyphotic-like postural stress on the sternum. Classification of the SSD based on chronicity suggests that SSD likely occurred within 2 time periods that coincide with more intensive phenotype testing periods of the pipeline (Fig. 1). However, the prevalence of SSD in an equivalent population size of 16-week-old C57BL/6N mice that did not go through the phenotyping pipeline is unknown, so this association can only be assumed. Overall, males were more frequently affected although the underlying reason for this bias was not determined. One possible explanation is males are heavier and may bear more weight on their sternum.¹¹

Relevance and Implications of Findings to Mouse Phenotyping and Welfare

Cardiac function in mice with RVEF was likely well preserved as the lesion was focal. We did not see any evidence of right ventricular cardiac insufficiency in any of the tissues examined. However, the lesion would likely alter the electrophysiological properties of the right ventricle such as conduction velocity. These changes have the potential to complicate or confound observations in cardiovascular studies, notably in aging mice. Replacement fibrosis of the right ventricle in senescent C57BL/6N mice was shown to underlie conduction slowing and arrhythmia.^17,27 Therefore we recommend that the sternum is routinely included in pathological evaluation of mutant mice, particularly in strains with or expected to have cardiovascular phenotype. Conversely, mice with thoracic skeletal abnormality such as kyphosis or generalized bone fragility should be screened for potential cardiac complications that may arise from underlying SSD. In the absence of comprehensive histopathology, such lesions might be interpreted as primary cardiac phenotypes. For the majority of cases, it is unlikely that the sternal lesions would be readily detected by X-ray imaging as was the case in this study. In humans, SSD is generally considered benign and treatment in most cases is limited to observation¹⁸ or conservative intervention by closed reduction.²³ Our observations in older SSD cases also suggest that this lesion is likely self-limiting in mice and resolves with proper reunion of the affected segments by bridging callus.

In summary, this case series study shows that sternal segment dislocation could occur in mice undergoing extensive and repetitive procedural work and that this lesion could result in right ventricular epicardial and subepicardial fibrosis. The underlying cause of sternal dislocation and the potential role of animal handling remain to be determined. However, awareness of this condition is an important first step to drive refresher training and assessment of handling practices during husbandry and procedural activities. Furthermore, it should prompt revision and refinement of the current tail lifting and neck-scruff restraining techniques that may cause transient kyphotic-like postural stress on the sternum. For instance, if operationally amenable, introduction of a scooping method or use of the home-cage tunnel to lift mice may be considered.¹⁰ Modifications to handling methods might help minimize the incidence of sternal dislocation and its potential confounding effect on data interpretation from phenotyping pipelines, notably those targeting the cardiovascular and skeletal systems.