The objective of this study was to monitor proprioceptors in the human pericardium. RLCs were the only corpuscles detected in substantial quantities. According to their position within the three pericardial sublayers, the RLCs were assigned to three localization classes (Fig. 3). We were able to verify the tripartite histological organization of the fibrous pericardium in all specimens studied in accordance with the literature [29, 30, 73]. Most RLCs were located between the lamina intermedia and lamina externa (class 2), and second most at the outer border of the pericardium within the lamina externa (class 3). We transferred the proprioceptors detected in the fibrous pericardial portions of all eight pericardia to macroscopic models (Fig. 5). Cluster analysis revealed the highest receptor density in the region of the ventricular bulges and at the outlet of the great vessels (Fig. 7). It can be inferred from these findings on pericardial propriosensing that displacements within the fibrous pericardial sublayers (class 1 and 2) and expansions at the outer pericardial border to the environment (class 3) are registered. These relative movements are affected by cardiac parameters, such as heart rate, contraction/relaxation, volume shifts (end-systolic volume, end-diastolic volume), and ejection fraction. Besides, even higher level perceptions such as time awareness [3] could be deduced.

In the musculoskeletal system, Ruffini corpuscles primarily register stretching, i.e., displacements or movements of tissue layers relative to each other [9, 27]. However, nothing can be found on their function in the viscera. Because they have yet not been studied electrophysiologically in internal organs, their relevance remains unclear and can only be assumed by analogies [75]. To date, no mechanoreceptors have been studied directly and precisely in either animal or human pericardium. We detected most receptors between the lamina intermedia and lamina externa and within the lamina externa bordering the thoracic cavity. Consequently, local gliding is likely to be particularly important at these sites, whereas the small number of receptors found between the lamina interna and lamina intermedia probably reflects a subordinate role in relative movements of the pericardial sublayers. Furthermore, this arrangement allows distinguishing displacements between the outer fibrous pericardial sublayers, i.e., of the pericardium itself and displacements of the pericardium against the adjacent tissues. We suggest that Ruffini corpuscles primarily register displacements between pericardial sublayers and between the pericardium and its adjacent structures. Yet, interpretation of the importance of the localization classes is limited as detailed biomechanical studies on the significance of the three pericardial sublayers are lacking.

As documented in the literature [58], we confirmed the sympathetic efferent innervation (tyrosine hydroxylase) of the RLCs, presumably allowing for regulatory control of the response thresholds to adrenergic stimuli. The significance of autonomic innervation of RLCs in the musculoskeletal system is poorly understood. Therefore, no direct analogies can be drawn regarding lowering or raising thresholds under increased sympathetic tone. Nevertheless, threshold rise has been documented for sympathetic activation, lowering pain perception [14]. Overall, it is believed that RLCs are at least involved in pain perception [64]. This is confirmed by the presence of Ruffini corpuscles alongside nociceptive fibers [34, 45], suggesting a potential indirect role in modulating pain through proprioceptive feedback. However, in accordance with others, we could not demonstrate CGRP-positive fibers in RLCs [65], which are typically involved in pain perception [32].

Kostreva and Pontus [37] also reported mechanoreceptor conduction in the canine pericardium. However, unlike our findings in the human pericardium, these receptors are concentrated along the atria and atrioventricular grooves. We observed that receptors were concentrated in the area of the ventricular bulges and at the outlet of the great vessels. Evolutionarily, this discrepancy may be attributed to the bipedal walk. Human’s erect posture causes a significant amount of blood volume to enter the heart following the force of gravity and be ejected upward against the force of gravity. This results in different volume dynamics compared to pumping blood perpendicular to gravity. Furthermore, the human heart is more suspended on its great vessels due to upright gait, while in quadrupeds, it resides primarily on its superior or lateral ventricular walls. Besides, Lee and Boughner [41] demonstrated that there are discrepancies in the thickness and viscoelastic properties of the human and canine pericardium. Consequently, the animal model's validity for humans is restricted.

The knowledge regarding the mechanoreceptors in the pericardium of animals is notably limited. Morphological and electrophysiological studies have identified nerve fibers or plexuses in a few animals, which suggests a mechanoreceptive function of the pericardium [18, 37, 61]. Some studies have investigated the potential role of mechanosensors in the pericardium and the activities of proprioceptive neurons in the spinal cord and cortical regions in processing mechanoreceptive information [43]. However, as we only identified studies that were likely derived from RLC-equipped pericardia of dogs or guinea pigs, it has to be stated that there are not sufficient morphological studies to improve functional understanding by comparative anatomy or evolutionary analyses. Nevertheless, the impact of these control circuits on the afferent side on heart dysfunctions offers interesting possibilities for therapy.

Our cluster analysis revealed four potential partitions and four hotspots of high density among the pericardial proprioceptors (Fig. 7). The clustering of pericardial receptor partitions corresponds to the heart's compartmentalization into four distinct chambers (in Fig. 7A, B, blue = right atrium, red = right ventricle, green = left ventricle, yellow = left atrium). Cluster boundaries are located at the anatomical borders of the heart between the four chambers. The interventricular septum separates the two ventricles (red and green cluster), while the valvular plane isolates the atria from the corresponding ventricles (blue and red cluster and green and yellow cluster). Although we see four separate measurement zones, it can be assumed that shifts in the pericardium at one location lead to signaling at the others.

Four pericardial proprioceptor hotspots (Fig. 7C, E) are exposed using the ward function. A prominent cluster of RLCs is located at both ventricular bulges, where the radius perpendicular to the longitudinal axis of the heart is the widest, indicating the outermost extension of the ventricles. Additionally, receptors accumulate at the outlet of the great vessels, particularly around the ascending aorta, where the visceral sheet of the serous pericardium transitions to the parietal sheet. These regions undergo significant volume changes during the cardiac cycle. Pronounced volume-related stretching of the pericardial layers is, therefore, sensed by a particularly large number of RLCs, emphasizing the most robust signals. For the hotspot surrounding the outlet of the great vessels, the fluctuation in diameter of the aorta and the pulmonary trunk (i.e., Windkessel function) must be considered. Volume fluctuations during swallowing may explain the additional receptor hotspot in the region of the left atrium bordering the esophagus. The low receptor density in the region of the interventricular septum may be attributed to its extraordinary role in the biomechanics of the cardiac cycle [68], whereas the missing receptors in the apex may be due to its limited exposure to volumetric changes, as it remains more stationary during the cardiac actions of the ventricles [28]. Accordingly, monitoring cardiac volume changes through pericardial proprioceptors does not appear relevant at these sites.

Sherrington [60] described proprioceptive “reflex arcs”, in which a motor response was observed as a feedback mechanism to a proprioceptive stimulus. Tuthill and Azim [69] reviewed the existence of supraspinal ascending proprioceptive pathways, in addition to the basic spinal reflex pathways that directly signal from proprioceptors to interneurons and motoneurons in the spinal cord. Ruffini corpuscles might serve as an external control mechanism capable of perceiving the stretching of the heart during cardiac cycle. However, proprioceptors in the human pericardium cannot provide the same level of control arc as extrinsic reflexes of skeletal muscles, as the heart lacks direct neuronal motor excitation. Nonetheless, proprioceptors likely modulate various cardiac parameters. It is generally accepted that the autonomic nervous system has an influence on cardiac output [11, 54]. RLCs detect shifts between pericardial sublayers and could be relevant for cardiac feedback by modulating the sympathetic or parasympathetic nervous system. Proprioceptive monitoring may, therefore, lead to changes in dromotropy, lusitropy, chronotropy, inotropy, and bathmotropy, either individually or in any combination. However, relative to each other, shifts of the pericardial sublayers depend on volume changes in time, location, and rate. As a result, Ruffini corpuscles would primarily provide information on heart rate (chronotropy) and stroke volume (inotropy and lusitropy).

Potentially disturbed pericardial feedback could play a critical role in the pathogenesis of several diseases. Changes in movement and distension of pericardial regions may cause significant alterations in sensory signaling. This could, for example, be attributed to an increase in the amount of pericardial fluid, such as in cases of pericardial serous effusion or hemorrhage. In such instances, the rapidly developing hemorrhage leads to greater receptor excitement resulting from the fast displacement of pericardial sublayers than it would occur during a slower increase in serous fluid production. On the other hand, an existing effusion, by utilizing the stretching reserve combined with a smaller extension of the heart, leads to fewer displacements of the pericardial layers. This will likely result in less contrasted feedback, negatively impacting cardiac regulation. The subsequent inadequate proprioceptive feedback may contribute to palpitations.

Fibrotic remodeling of the pericardium following pericarditis or intervention-induced scar formation reduces elasticity, causing the pericardial sublayers to adhere to one another. This adhesion limits or entirely prevents the expansion and relative sliding of the pericardial sublayers. Consequently, the afferent limb of the pericardial loop becomes inadequate.

Disrupting the pericardium’s integrity, such as in pericardial opening during surgical procedures or pericardiectomy, may lead to decreased sensitivity. Particularly pericardiectomy, implemented as a treatment for pericarditis, is associated with high perioperative and long-term mortality rates [4, 20, 46]. This may indicate that the pericardium performs mechanical functions and is also involved in feedback to monitor cardiac function. A resection of the pericardium or cases where the pericardium is left open could lead to a partial or complete loss of the afferent proprioceptive pericardial limb. This might be associated with life-threatening symptoms.

It is relevant to note that in the context of pericardial diseases in the scientific literature, the pathophysiology is mainly reduced to mechanical consequences, and others, e.g. sensory functions, are not taken into account. This may be due to the lack of information about sensory feedback from the pericardium. However, it is crucial to consider impaired regulation as another relevant factor in the pathophysiology of constrictive pericarditis, cardiac tamponade, and pericardial integrity disturbances. This perspective also suggests potential avenues for therapeutic interventions [2].

The use of pericardial tissue exclusively from body donors, with an average age of 82 years, led to limitations of this study. Potentially, the number and distribution of RLCs may differ in younger age cohorts and change with aging. It has been observed in both human and animal musculoskeletal systems that the number of proprioceptors tends to decline with age [5, 42, 44]. However, Ruffini corpuscles have a unique capacity for regeneration [77]. Therefore, objective evaluations of the receptor number in younger populations and their possible dynamics cannot be conducted. Furthermore, only RLCs can be accurately identified when their fusiform shape is cut longitudinally parallel to the orientation of the collagen. It is plausible that there may exist corpuscles running vertically between the pericardial sublayers, but their identification remains uncertain. Since the extent of this component is unclear, we have refrained from extrapolating the total amount of RLCs to the entire pericardial surface. Despite the limited number of pericardia, we observed clear foci in the receptor distribution that were consistent throughout all specimens. Because of the limited comparable literature, there is no further support for explaining this striking spatial distribution. These observations thus serve as a foundation for following functional studies and provide a suitable basis for digital experiments.

Our study demonstrated the existence of RLCs within the fibrous pericardium. Confirmation with other methods, such as electrophysiological studies or evidence of feedback mechanisms, would be required for further evidence. However, the high number of receptors suggests significant implications. Therefore, we suggest that pericardium serves an undiscovered function as a sensor with the RLCs as its anatomical correlate.

(1) RLCs are present in large numbers in the human pericardium, (2) Based on the tripartite histological organization of the human pericardium, they can be classified into three localization groups, (3) The receptor distribution follows a distinct pattern with highest density in the region of the largest ventricular diameter and at the transition to the large vessels, and (4) No sex-specific differences in the number or distribution of receptors can be detected.

Inferential, the human pericardium is subject to neural proprioceptive control.