An extravascular fluid transport system based on structural framework of fibrous connective tissues in human body

Abstract Objective Interstitial fluid in extracellular matrices may not be totally fixed but partially flow through long‐distance oriented fibrous connective tissues via physical mechanisms. We hypothesized there is a long‐distance interstitial fluid transport network beyond vascular circulations. Materials and methods We first used 20 volunteers to determine hypodermic entrant points to visualize long‐distance extravascular pathway by MRI. We then investigated the extravascular pathways initiating from the point of thumb in cadavers by chest compressor. The distributions and structures of long‐distance pathways from extremity ending to associated visceral structures were identified. Results Using fluorescent tracer, the pathways from right thumb to right atrium wall near chest were visualized in seven of 10 subjects. The cutaneous pathways were found in dermic, hypodermic and fascial tissues of hand and forearm. The perivascular pathways were along the veins of arm, axillary sheath, superior vena cava and into the superficial tissues on right atrium. Histological and micro‐CT data showed these pathways were neither blood nor lymphatic vessels but long‐distance oriented fibrous matrices, which contained the longitudinally assembled micro‐scale fibres consistently from thumb to superficial tissues on right atrium. Conclusions These data revealed the structural framework of the fibrous extracellular matrices in oriented fibrous connective tissues was of the long‐distance assembled fibres throughout human body. Along fibres, interstitial fluid can systemically transport by certain driving‐transfer mechanisms beyond vascular circulations.


| INTRODUC TI ON
The earliest record of the long-distance extravascular fluid flow was the perivascular spaces in brain around 1850s. However, fluid in the interstitial matrices is thought to be mainly entrapped locally by proteoglycan filaments 1 and local fluid in the distal end of human extremities is retaken only by venous and lymphatic vessels. Whether fluid in the interstitial matrix can flow systemically throughout the body beyond vascular circulations has been debated for decades. 2,3 Recently, several studies showed that partial interstitial fluid can flow along tunica adventitia or through a macroscopic fluid-filled interstitial space, forming an extravascular transport pathway for interstitial fluid. [2][3][4] Starting from 2006, we focused on a long-distance transport phenomena of interstitial fluid in animal and humans and found two types of long-distance extravascular fluid transport pathways, a cutaneous and a perivascular pathway. 4,5 For example, in rabbit, extravascular fluid transport from the lower limb into pericardial cavity was found to be via the venous adventitia and its surrounding fibrous connective tissues. 4 In one amputee of lower leg patient, who received the hypodermic injection of the fluorescent tracer into ankle dermis before taking an amputation of lower leg due to severe gangrene of foot, extravascular fluid transport from ankle dermis to the broken end of lower leg was anatomically identified to be the cutaneous pathway (located in dermis and the interlobular septum among hypodermic adipose tissues) and the perivascular pathway (located in general connective tissues surrounding the veins and the arteries), and both pathways were composed of the oriented fibrous connective tissues. 5 By twophoton confocal laser microscopy (TPCLSM), the longitudinal fibres along the transport long-axis of these pathways were stained by the fluorescein from ankle. 5 However, fibrous interstitial matrices are usually considered to be a three-dimensional architecture with the isotropic fibres. The reason that why only the fibres towards the transport long-axis were stained by the fluorescein was unclear.
In healthy volunteers, two peripheral long-distance extravascular fluid transport pathways originating from acupuncture points in the extremity endings were also found by MRI. [5][6][7] Our experiments showed there were two types of "the entrant points or hypodermic entrances" for the tracer injection to visualize the long-distance extravascular pathways of the extremities. 5,6 One was a hypodermic point in the vicinity of venous vessels of the forearm or lower legs, which was used to visualize the perivascular pathways. The other was a hypodermic point at the distal end of the fingers or toes or around the wrist or ankle, which was usually an acupuncture point and applied to visualize both the cutaneous and perivascular pathways. However, the comprehensive distributions of the long-distance extravascular pathways have not been fully understood due to the unfitness of current contrast-enhanced MRI technique in tracing the paramagnetic agent in the visceral organs of volunteers. The connections of the peripheral extravascular pathways with the associated visceral structures were still unknown in human.
Moreover, both the kinetic/dynamic mechanism and regulating factors for these long-distance extravascular fluid transport through fibrous matrices remain unknown. For fibrous matrix, neither the collagenous and elastic fibres nor the gel-like substances can flow freely. 1 The demonstrated staining processes of the fluorescent fluid along the fibres suggested that an interfacial transport zone between the solid fibre and the surrounding gel-like substances could be responsible for the fluid transport. 5 As to regulating factors, our pilot study in the amputated legs revealed that the mechanical force is one of dynamic origins to "pull or push" fluid from the extravascular pathways. 5 Our study in rabbits showed the peripheral extravascular fluid from ankle was transported along the perivenous fibrous tissues surrounding the inferior vena cava, three grooves of the heart and into the pericardial cavity to form pericardial fluid. 4 Thus, we speculate the mechanical heart beatings might be one of the regulating factors to drive fluid through the oriented fibrous matrices of connective tissues in physiological conditions.
In current study, we first used twenty healthy volunteers to further demonstrate two types of the hypodermic entrant points by MRI where are acupuncture points in the extremity endings. We then chose one of the entrances identified in healthy volunteers as the hypodermic injection point in human cadavers. To visualize the long-distance extravascular pathways, we used a mechanical chest compressor device to simulate the heart beatings to drive interstitial fluid flow. 8 The anatomic distributions and histological structures of extravascular pathways from the extremity ending to the associated visceral structures were consequently identified. By the current experiment of dynamic gross anatomy, we may verify the structural framework of the oriented long-distance extravascular fluid transport pathways throughout human body and the hypothesis there is a longdistance interstitial fluid transport network composed of oriented fibrous matrices of connective tissues beyond vascular circulations.

| MRI image of the pathways from acupuncture points of hand and foot in volunteers
Twenty healthy volunteers aged from 23 to 37 years (15 men and five women) were recruited. Exclusion criteria included a history of primary or secondary extremity lymphedema, obesity (lipoedema), Conclusions: These data revealed the structural framework of the fibrous extracellular matrices in oriented fibrous connective tissues was of the long-distance assembled fibres throughout human body. Along fibres, interstitial fluid can systemically transport by certain driving-transfer mechanisms beyond vascular circulations. extensive scarring or dermatological abnormalities in the areas tested and contraindications for MRI. The study was approved by the ethics committee of Beijing Hospital (No. 2016BJYYEC-066-02).
All the participants have signed the informed consent prior to the initiation of the study.
Five Jing-well points on fingers, an acupuncture point on wrist, three Jing-well points on toes and three acupuncture points on ankle were selected as the hypodermic injection points for MRI. Details of injection points, the locations and the corresponding acupuncture points were listed in Table 1 and Figure 1.
Each injection point was tested on two volunteers. The choice of right or left hand or foot depended on the will and preference of the volunteers. Two volunteers received the routine MR angiography on forearm. After the angiography, an injection point, which was near a superficial vein and neither acupuncture point nor a point on Meridians, was selected as the control. The depth of the injection was approximately 1-2 mm into subcutaneous loose connective tissues of the point. The paramagnetic contrast agent, gadolinium diethylenetriamine penta-acetic acid (Gd-DTPA, Magnevist; Bayer Schering Pharma AG), was hypodermically administered at minimal volumes of 0.2-0.5 mL (the final diluted concentration was 156 mg/mL). The transport pathways originating from the acupuncture points were scanned by an Achieva 3.0T TX scanner (Philips Electronics) or 1.5T MR Scanner (GE Medical Systems). The images were obtained using a 3-dimensional T1-weighted spoiled gradient-echo sequence (SPGR) or fast field echo (FFE) sequence using an 8-channel phased-array head coil. Scanning parameters were adjusted to obtain increased spatial resolution. The raw data were analysed using an extended MR WorkSpace station with maximum intensity projection reconstructions (MIPs).

| General medical information of the donated human cadavers
The human cadaver experiment was supported by the Human

| Hypodermic injection on the right thumb and the visualization of the long-distance extravascular fluid transport pathway in cadavers
To visualize the long-distance fluid transport pathway, the distal end of right thumb was selected as the injection point in this study.

| Visualization of the lymphatic vessels on hand and forearm
The other two cadaver subjects (No. 14, 15) were given pressure injection with Chinese ink into the same injection point of right thumb as first two groups to visualize lymphatic vessels on hand and forearm. The dissection processes were recorded by digital camera.

| Extravascular fluid transport pathways visualized by quantum dot microspheres
The two cadaver subjects (No, 16,17) were given the injection of

| Histological study and two-photon laser scanning microscopy examination on the fluorescent pathways
The fluorescently stained tissues on hand, forearm, upper arm, axilla and thorax were sampled, respectively. Sampled tissues were

| Immunohistochemical staining on the fluorescent pathways
Immunostaining was performed according to standard procedures.
The epitope retrieval was performed manually by using pressure cooker. Envision + HRP (Dakocytomation) was subsequently used as secondary detection. The samples were frozen-dried for 24 hours and immersed into 10 wt% silver nitrate solution for 30 minutes. 3 wt% sodium hydroxide solution was dropped in slowly with the same volume used for silver nitrate. Half an hour later, the samples were gently washed by deionized water and frozen-dried again. The prepared samples were preserved in room temperature for the micro-CT scanning.

| X-ray 3D micro-CT on the fluorescent pathways
The samples were imaged using a high-resolution 3D X-ray microcomputed tomography (nanoVoxel-3000 series; Sanying Precision Instruments Co., Ltd.) according to standard procedures. The parameters were adjusted to obtain higher resolutions.
The images were captured on a detector 1920 × 1536 pixels, with an exposure time of 0.6 seconds and a voxel size of 4.6 μm approximately.

| The cutaneous pathways and the perivascular pathways from the hypodermic entrances identified in 20 volunteers by MRI
When the injection points were acupuncture points (Table 1,

| The fluorescent pathways were neither blood nor lymphatic vessels verified by gross anatomic methods
The long-distance fluorescent pathways were neither blood nor lymphatic vessels. Firstly, the fluorescent pathways were not the re- All these were not tubular structures.

| Histological analysis on the fluorescent pathways
Samples were taken from the skin of the hand and forearm, the ce- However, fluorescent fluid through arterial adventitia has not been seen here.

| Three-dimensional structures of the intrinsic fibre network of the fluorescent pathways observed by micro-CT
By micro-CT method, 9 the intrinsic three-dimensional architecture of the fluorescent pathways was clearly displayed (Figure 5C,D). The

| D ISCUSS I ON
Beyond vascular circulations, we identified a cutaneous and perivascular fluid transport pathway that was composed of fibrous matrices of oriented connective tissues in human body. In line with our previous findings in the extremities, 6,7 the anatomic connection of these extravascular pathways from the distal end of thumb to the right atrium disclosed a profile of the systemic distributions of the long-distance interstitial fluid transport pathway network. Our MRI results in volunteers indicated that more long-distance extravascular pathways were initiated from the hypodermic entrances of other fingers and toes, usually from acupuncture points. Thus, F I G U R E 5 The intrinsic fibres of fibrous connective tissues in different anatomic sites disclosed by micro-CT and TPLSM. A, two cutaneous pathways on skin. B, sampled between the pathways and showed irregular interlobular septum in hypodermic tissues. C, sampled from the cutaneous pathway and showed longitudinal interlobular septum towards the transport direction. D, scanned in the middle of (C) and showed the internal fibres of septum were distributed mainly longitudinally towards the transport direction. E, the fluorescently stained fibres within the cutaneous pathways in comparison with (F) (sampled from the left forearm as control without fluorescein transport). H and I, the 3D structures of the adventitia of (G). I, one plane in the yellow region of (H) and showed the longitudinal fibres were enriched in one side of the "conduit wall." J, the fluorescently stained fibres within the adventitia of (G) in comparison with (K) (sampled from the left axillary vein as control without fluorescein transport). L, the fluorescently stained superficial tissues on the right atrial wall and pectinated muscles of atrial appendage. M, the intrinsic structures of the right atrial appendage and the superficial tissues (yellow box). N, the criss-crossed fibres in the superficial tissues on pectinated muscles by micro-CT. O, the fluorescently stained fibres within the right atrial appendage in comparison with (P) (sampled from another cadaver's right atrial appendage without fluorescein transport as control) their connections with the associated visceral organs or the structures need further verification in both the cadavers and volunteers.
Considering that fibrous connective tissues are one of the four basic tissues and distributed in almost everywhere, we proposed that fluid in fibrous interstitial matrices is not totally fixed in tissue gel but transport partially along the extravascular pathways of oriented fibrous connective tissues under certain dynamic physical mechanisms, forming a network of long-distance interstitial fluid transport F I G U R E 6 Illustration of the fluorescent pathways from LU11 of the 1 cadaver F/92. By the simulated heart beatings, the fluorescein from the fingertip visualized the fluorescent pathways that divided into brachial plexus on hand (A1) and wrist (A5). Interestingly, a branch went centrifugally to the bottom of the index and middle finger (A1, A2). The perivascular pathway was sometimes enveloped in the cutaneous pathway (A2, A3) or beside it (A6). The cutaneous pathways could contain diverse anatomic structures in one plane, including the dermis, hypodermic tissues, superficial fascia, deep fascia on the tendons of extensor pollicis longus, extensor pollicis brevis and pollicis longus (A4). B1, the hypodermic tissues under natural light of the fluorescently stained tissues of (B2). C1 and C2, the outer wall of superior vena cava, and right atrium was stained by the fluorescein F I G U R E 7 Illustration of 10 nm quantum dot microsphere on a venous vessel. After 2.5 h of compression on the heart, the 10 nm quantum dot microspheres from the right thumb can be found on a venous vessel in the forearm of the No. 16 subject pathways systemically throughout the human body or other animal bodies.
The long-distance perivascular fluid transport was reported in several parts of the body, such as the perivascular spaces and the glymphatic pathways along the periarterial sheaths and the largecaliber draining veins of mice brain, 2,10-12 the subcutaneous "perivascular-spaces-like" channel near the lymphatic and blood vessels of the abdominal skin of rabbits, 13 the perivenous pathways along the adventitia of the lower extremity veins, inferior vena cava and into the pericardial cavity, 4 some segments of small intestines and partial pulmonary veins of rabbits, 4 and the perivenous pathways and the periarterial pathways in human lower legs. 5 In fact, not only fibrous adventitia but also its surrounding general loose connective tissues were confirmed to be the perivascular pathways for fluid flow. 5 These studies indicated that fluid around the entire vascular tree, including both venous and arterial adventitia, would be transported along fibrous adventitia and converge into the superficial tissues of the right atrium or other parts of the heart in physiological conditions. The revealed function of fibrous adventitia may trigger new inspiration on vascular biology ( Figures 3A,B,C and 6B1,B2).
The long-distance cutaneous pathways were also histologically identified as oriented fibrous matrices of connective tissues.
Unlike vascular tree connects with the heart, the panorama distributions of the cutaneous pathways have not been understood yet. The cutaneous pathway from the distal end of thumb formed a plexus in hand and forearm ( Figures 3I and 6A1,A5), which were composed of the millimetric interlobular septum of adipose tissues longitudinally assembled towards the transport long-axis (Figures 3D,F and 5C,D; Video S1). By contrast, the dermic tissue beyond the cutaneous pathways was of irregular interlobular septum ( Figures 3E,F and   5B; Video S3). Above the elbow level of the cadavers, the cutaneous pathways were not found and seemed to converge into the perivascular pathways along the cephalic and basilic vein, axillary sheath and the superficial tissues of right atrium eventually ( Figures 3C,B,J, 5G,L and 6B1-2, C1-2). Whereas in volunteers by MRI, the cutaneous pathways from hand were found in the entire upper limb. 6 The anatomic approach may not disclose the panorama distributions of the cutaneous pathways in volunteers.
The reasons may be due to the dynamic factors to drive interstitial fluid flow. In current anatomic study, only parts of the heart near the chest were motivated by the device. Both the cutaneous pathways in the upper arm and the periarterial pathways have not been observed in the cadavers. In physiological conditions, we have verified in one amputated lower leg patient, who took the tracer injection before the amputation, that there were four types of longdistance extravascular fluid transport pathways at least: a cutaneous pathway, a perivenous pathway, a periarterial pathway and even a fibrous-endoneurium-perineurium-epineurium pathway. 5 Thus, it indicated there would be more extravascular fluid pathways driven by more dynamic factors in physiological conditions. Specific medical imaging technique was needed in alive human body to visualize the diverse long-distance extravascular fluid transport pathways under the guidance of the disclosed anatomic structures in cadavers.

| The intrinsic structures of the long-distance extravascular fluid transport pathways with oriented fibrous matrices of connective tissues
The fibrous matrix of oriented connective tissues was identified to be the histological structures of the long-distance interstitial fluid transport pathways as well as the same histological tissues of the interstitial matrix of extracellular matrices, fibrous cellular environment, perivascular spaces in brain, paravenous pathways, perivascular sheath, fibrous sheath, nerve sheath, epineurium, endoneurium, fascia, etc. Until now, the intrinsic three-dimensional architectures of fibrous matrices and their long-distance interconnections systemically throughout the human body have not been fully understood yet.
Revealed by micro-CT, the millimetric tunica adventitia and longitudinal interlobular septum were identified to be composed of the micron-sized fibres which were mainly oriented towards the transport long-axis. By histological analysis, the long-distance ex-

| An interfacial transport zone between the solid fibres and the surrounding gel-like substances is responsible for fluid transport through the longdistance fibrous extravascular pathways
Because the fibres and the gel-like substances are fixed and cannot flow freely, 17,18 the only possible free space that allows fluid to transport is the interfacial clearances between fibres and gel-like substances, which may be termed as "fiber/gel interfacial clearance" and abbreviated as "interfacial clearance." Once the fluid is filled into the interfacial clearance, then a liquid film may be formed, which may be named as the "fiber/gel interfacial liquid zone" and abbreviated as "interfacial liquid zone." Once the liquid flows in the zone along fibres under certain dynamic driving mechanism, then an "interfacial transport zone" may be set up. 19,20 Because the fibres are assembled longitudinally through a long-distance fibrous extravascular pathway, the interfacial transport zone is of long-distance characteristics. Besides, along the fibrous framework of oriented connective tissues, the long-distance interfacial transport zones may be topologically connected and the "network of interfacial transport zones" in fibrous matrices of connective tissues throughout the whole body is created.
By using TPLSM, the abundant fluorescently stained fibres along the transport long-axis were observed and the results of the fluorescent fluid transport through the long-distance fibrous extravascular pathways, which clearly represented an imaging characteristic for the interfacial transport zones in fibrous matrices ( Figure 5E,J,O).
The network of interfacial transport zones may act as interstitial fluid transport pathways in the meshwork of fibrous connective tissues throughout the whole body. 19 The present data were in agreement with our previous studies in the amputated lower legs. 5 Illustrated by the 10 and 110 nm quantum dot microspheres, the pore sizes of the interfacial transport zones of the perivenous pathways may be <110 nm in the involved cadavers (Figure 7). The exact pore sizes of the interfacial transport zones of diverse fibrous connective tissues in physiological conditions need further studies.

| The dynamic transport mechanism for the systemic extravascular fluid transport network
For the widespread fibrous matrices of connective tissues over the whole body, the dynamic transport mechanisms and regulating factors for the network of the long-distance extravascular fluid transport have not been comprehensively studied. In the studies on the glymphatic system in brain, fluid transport through the perivascular pathways is related to the cerebrovascular pulsatility, body posture and perivascular astrocytic aquaporin-4. [10][11][12] In previous study on rabbits, fluid flow along venous adventitia into pericardial cavity was affected by the intactness and the wettability of the fibrous framework of the perivascular pathways as well as the periodic heart movements. 4 As for the dynamic pattern of the long-distance extravascular fluid transport pathways driven by the periodic to-and-fro mechanical movements of right atrium in this dynamic gross anatomic study, the following multiscale dynamic driving mechanisms from macroscopic scales to microscopic scales may be proposed. The repeated compressions of the mechanical device on chest generated a macroscopic driving centre, which was the right atrial wall and appendage near the chest here. At the microscopic scale, the long-distance interfacial transport zones of fibrous extravascular pathways from the extremity ending to the driving centre provided a topologically connected interfacial clearances for interfacial liquid film throughout the oriented fibrous connective tissues. The periodic to-and-fro mechanical movements of the superficial tissues of the right atrium would act on the gel-like fibrous tissues of the driving centre and lead to the periodical fluctuation of pressure. The periodical fluctuation of pressure will force the gel-like tissues to compress and relax periodically, like a pump. The periodical compression-relaxation of the pump played a role to "pull" the topologically connected interfacial liquid film in a long-distance extravascular pathway. The exact roles of "driving-center" on "interfacial-liquid-film-transport" were complicated. We hypothesized that water molecules in interfacial liquid film would assemble a long-distance link-chain of hydrogen-bonding or Van der waals' interactions to form "gearing-chains," which together with interfacial interactions may allow the long-distance interfacial liquid films to be "pulled" towards a driving centre. 19,[21][22][23] More experiments are necessary to verify the hypothesis of multiscale dynamic driving mechanisms, which is "fibro-tissue driving-transfer mechanisms for fluid transport," named as "gel pump." As for the relationship between the long-distance fibrous extravascular pathways and the vascular circulations, we verified in alive rabbits that fluid in ankle dermis would converge into the perivascular pathways, the venous and lymphatic vessels in the meanwhile. 4 As fibrous matrix of connective tissues is the bed for capillaries, free interstitial fluid that was transported via the long-distance fibrous extravascular pathways may easily enter capillary vessels by Starling forces in physiological conditions. In the current anatomic study, the intraluminal blood taken from right ventricular cavity and the opposite arm has not been fluorescently stained during the whole experiments, which indicated the long-distance fibrous extravascular pathways were relatively independent from vascular circulation.
In alive human body, the exact relation of the long-distance fibrous extravascular pathways with vascular pulsation, Starling forces, interstitial pressure gradients, the electrical activities of nerves, the temperature and even electromagnetic force need further explorations. 18

| A potential relationship between the systemic fibrous extravascular pathways network and Traditional Meridians and Collaterals network
In accordance with previous findings in volunteers by MRI, 6  In summary, fibrous matrices of connective tissues are thought to contain irregular three-dimensional matrices and act as glue for cells to attach to matrix. Our data verified the structural framework of an oriented fibrous connective tissue was composed of the multi-layered, longitudinally assembled and intertwined micron-sized fibres from an extremity ending to an associated visceral structure. Upon the network of topologically connected interfacial transport zones along the long-distance, oriented and orderly assembled fibres, interstitial fluid can be transported in the meshwork of fibrous connective tissues, named as "fibro-tissue interfacial transport." These data strongly suggest that interstitial fluids in the extracellular matrices are not fixed in tissue gel but systemically transport throughout human body and enter blood vessels and initial lymphatic vessels eventually by the Starling forces. We hypothesized that: (a) the systemic interstitial fluid transport together with the cardiovascular system and lymphatic system comprises three principal components of circulatory system in animals and humans.

CO N FLI C T O F I NTE R E S T
The authors declare no competing financial interests and have no conflicts to disclose.

AUTH O R CO NTR I B UTI O N S
HyL conceived the conception and designed the experiments.
HyL, FsJ, FW and ZA analysed the micro-CT data. HyL, MC, WtL and HL analysed the MRI data. CM proved the usage of human cadavers.
HyL and YjY analysed the interfacial transport pattern. HyL wrote the manuscript. LjS made English corrections. All authors contributed to scientific discussions of the manuscript.

DATA ACCE SS I B I LIT Y
The data that support the findings of this study are available on request from the corresponding author. The data of the donated cadavers are not publicly available due to privacy and ethical restrictions.