INTRODUCTION
There is variation and some confusion in use of the term “paratenon” among clinicians. Benjamin et al. [1] provided concise definitions of the paratenon and its analogues as follows: the endotendon is a thin film of loose connective tissue providing a vascular route; the epitenon surrounds the tendon as a whole and is continuous with the endotendon; and the paratenon is a tendon sheath that has a different histology and is clearly separated from the tendon itself. These authors considered the paratenon of the Achilles tendon as the best example. Therefore, the paratenon is not usually continuous with the synovial lining of a joint; similarly, some bursas, such as the trochanteric bursa, do not communicate with the joint cavity.
The careful removal of the paratenon along the cruciate ligament is a critical step in knee surgery [3], and paratenon stripping is an important procedure in interventional radiology of the knee [14]. The paratenon of the knee refers specifically to a sheath of the ligament, in contrast to the other or more common use of this term to refer to a sheath of the Achilles tendon [8, 18, 19]. Clinicians commonly use the term “paratenonitis” to refer to inflammation of the outer layer of the tendon, regardless of whether the paratenon is lined by a synovium [12]. However, for knee surgeons and interventional radiologists, the paratenon is a basic anatomical tissue along a ligament, not along a tendon. In the knee joint, the synovial lining should cover the paratenon along the cruciate ligaments. Because the infrapatellar fat pad is the largest synovial plica (thin tissue fold) in the human body, it might communicate with the paratenon.
The general aim of this study was to describe the development and morphology of the paratenon along the cruciate ligaments of the knee in human foetuses. The cruciate ligaments develop in the interzone mesenchymal tissue, between the femur and tibia [5, 15, 16]. Thus, it is likely that the paratenon also develops from primary interzone tissue, rather than being divided from the established cruciate ligament. The specific aim of this study was to determine whether a synovial lining was present at the expanding joint cavity along and around the paratenon of the knee. Previous research showed that the cytoplasm-rich cells in foetal synovial tissues had a cuboidal appearance [9, 13].
MATERIALS AND METHODS
This study was performed in accordance with the provisions of the 2013 revision of the Declaration of Helsinki. For macroscopic observations, the left knees of 25 near-term foetuses (crow-rump length [CRL]: 272–310 mm; gestational age [GA]: 32–37 weeks) were examined. The patellar tendon was cut at the tibial attachment to reflect the patella and infrapatellar fat pad upward. During the reflection, an attachment of the fat pad to the femur was cut (if present). A slight dissection was performed along the cruciate ligaments, and photos were taken at each step using a Pentax K-1 camera with a 50–100 mm zoom lens.
For histological observations, sagittal sections of the unilateral knee from 8 midterm foetuses (CRL: 35–73 mm; GA: 8–12 weeks) and 10 near-term foetuses (CRL: 280–320 mm; GA: 33–39 weeks) were examined. The latter 10 specimens were not used for the macroscopic observations. All sections were stained with haematoxylin and eosin (H&E). Most histological photographs were taken using a Nikon Eclipse 80, but photographs at ultra-low magnification (objective lens less than ×1) were taken using a high-grade flat scanner with translucent illumination (Epson scanner GTX970).
The histological sections from 6 midterm foetuses were part of the large collection at the Department of Anatomy of the Universidad Complutense, Madrid, and were from miscarriages or ectopic pregnancies at the Department of Obstetrics of the University. Until 1980, the sections had been prepared and stained with H&E, Azan or silver impregnation. The used of these specimens for research was approved by the Ethics Committee of Complutense University (B08/374). The knees of near-term foetuses were part of the collection of the Department of Anatomy, Akita University, Japan and were donated by families to the Department in 1975–1985 and preserved in 10% w/w neutral formalin solution for more than 30 years. Data on these specimens included the date of donation and the number of gestational weeks, but not the name of the family, obstetrician, hospital, or reason for abortion. The use of these specimens for research was approved by the Akita University Ethics Committee (No. 1428).
The unilateral knees from the 10 near-term foetuses were divided, and the specimens were incubated at room temperature in Plank-Rychlo solution (AlCl2/6H2O, 7.0 w/v%; HCl, 3.6%; HCOOH, 4.6%) for 1 to 2 weeks. Routine procedures for paraffin-embedded histology were performed, and semiserial sections of knees were prepared at 0.1 mm intervals.
RESULTS
Macroscopic observations
When we removed the patella and patellar tendon to open the knee joint from the anterior side, we also pulled off the infrapatellar fat pad from the joint cavity because it was attached to the patellar tendon at the inferior side of the patella. However, because a marginal part of the fat pad was attached to the femur, we cut it before taking photos (Fig. 1).
The cruciate ligaments consistently had a smooth surface (Fig. 1A, C, E). However, this surface was not usually of the ligament itself, but of the paratenon. Twenty-three of the 25 knees examined macroscopically had anterior cruciate ligaments (ACLs) that were entirely covered by capsule-like paratenons (Fig. 1B), and these sometimes contained feather-like structures attached to the anterior aspect of the ACL (Fig. 1D). Thus, in contrast to the first impression upon opening of the knee, the actual ACL was rarely exposed to the joint cavity (Fig. 1E). The ACL was evident in the anterior view in the other two knees (Fig. 1E, F). When the tibia was twisted, a crossing between the ACL and the posterior cruciate ligament (PCL) was released, and provided an almost parallel arrangement (Fig. 1G). This posture allowed us to determine whether a residual paratenon was present or absent after dissection.
The paratenon also covered the PCL, but was usually much thinner there than along the ACL. In 6 of 25 knees in which the paratenon was thick along the PCL, it had the appearance of another ligament that merged with the ACL at the tibial attachment (i.e. the ligamentum mucosum, Fig. 1F). Therefore, in these 6 knees, the paratenon of the PCL tended to be thicker on the lower side due to the antero-inferior course of the ligamentum mucosum. We found no correlation between the foetal age and gross morphology of the paratenon.
Histological observations
At midterm, an interzone mesenchymal tissue between the femur and tibia gradually disappeared (Fig. 2A). However, at 9 weeks GA, most parts of the ACL were embedded in the interzone tissue remnant (Fig. 2B). At that time, a joint cavitation advanced into the posterolateral part. Thus, along the PCL, the interzone tissue was limited to near the femoral attachment (Fig. 2C). Development of the infrapatellar fat pad was much delayed relative to the cruciate ligaments (Fig. 2B).
At near-term, we found various stages of paratenon development, and no clear correlation of stage with foetus age or size. Thus, the larger knees did not usually have a sheath- or armour-like paratenon, and also did not usually have a larger joint cavity. A candidate initial stage of the paratenon differentiation was present in the posterior side of the ACL (Fig. 3A, B). In this case, the enlarging joint cavity was surrounded by synovial plica-like tissues and by ligament-like tight tissues; the latter was apparently divided from the ACL with long tears (Fig. 3B). In contrast, an initial change was limited to an interzone tissue remnant that extended inferiorly along the PCL (Fig. 3C, D).
At the next stage (CRL 320 mm), the ACL accompanied a fatty tissue-rich paratenon or synovial plica around it (Fig. 4) or was embedded in a venous plexus-like paratenon (Fig. 6). These loose tissues contained abundant veins and several arteries. In contrast, the PCL accompanied the paratenon at and near the bony attachments. The infrapatellar fat pad was thick and long, and composed of a mesh of fibrous bands (less than 0.1 mm thick) that contained fatty tissues between the bands (Fig. 4C). The fat pad surface and the paratenon surface to the joint cavity were lined by cuboidal cells (i.e. a primitive synovial lining, Fig. 5A, B, F). In contrast, flat cells lined a loose space along the ACL (Fig. 5C, D).
Notably, the cavity with flat lining cells continued to or communicated with the other non-lining spaces, which contained thin fibres (Figs. 4F and 5E) and/or fragmented tissues (Fig. 6C, D, G). Thus, at near-term, the drastic reconstruction of the paratenon was likely to occur with a cavitation along the ACL. Likewise, in the venous plexus-like paratenon (Fig. 7), vein fusion and degeneration provided a venous valve-like structure (i.e. a degenerating venous endothelium, Fig. 7B–E). This morphology was quite different from an actual vein that was surrounded by a flat endothelial lining (Fig. 7A). Therefore, the venous fusion was similar to the process of joint cavitation via fusion of small spaces in degenerative tissues (Fig. 5C). These processes of the cavitation and venous reconstruction in the paratenon exposed the nearby arteries to the joint cavity (Figs. 4F and 6G).
DISCUSSION
The present study is apparently the first histological report of the paratenon of the knee in human foetuses. Because these ligament-associated tissues are small and thin in midterm foetuses, we believe that the thick paratenons that we identified macroscopically in near-term foetuses (Fig. 1) likely developed from the interzone mesenchymal tissue between 13 and 31 weeks GA. In contrast to the development of a tight meniscus in the loose interzone tissue [4], a relatively loose paratenon did not have a definite shape during the prenatal period. However, in near-term foetuses, there was likely to be a drastic reconstruction of the paratenon, especially along the ACL. At this time, individual differences in the thickness of the paratenon became evident along the ACL, presumably due to differences in knee posture in utero. A slight twisting of the cruciate ligaments (i.e. a slight rotation at the knee) at this time probably pushed and interfered with enlargement of the paratenon. Actually, Katz et al. [10] reported greater external rotation of the tibia in babies after earlier delivery. We found two-types of histology: a fatty tissue-rich paratenon and a venous plexus-like paratenon. Because active cavitation in the paratenon was suggested in the venous plexus, it is unlikely that one type of paratenon could change into the other type. In both types, the paratenon contained abundant arteries and veins, especially at and near the crossing between the cruciate ligaments.
Notably, veins in the paratenon were likely to fuse and, because of limited blood supply from prenatal knee movements, the plexus seemed to degenerate. In the degeneration in near-term knees, the advancing fusion of venous cavities seemed to provide a large cavity without the usual cuboidal synovial cell lining. This might be a unique manner of joint cavitation, although it is somewhat similar to the fusion of thyroid follicles in the aged thyroid because of abundant valve-like fragments of the follicle wall [20]. A similar fusion is also seen in circadian changes of the follicles although it is followed by re-division: the size of follicle drastically changes [21]. The unique cavitation in the paratenon venous plexus exposed the nearby arteries to the joint cavity (Figs. 4F and 6G), and these arteries should be degenerated. In near-term foetuses, we recently reported a close topographical relationship between the tributaries of the external vertebral venous plexus and recesses of the zygapophysial joint cavity [13]. A degenerating venous plexus during joint cavitation might be a common method of development. Although we cannot deny the possibility that the present venous plexus-like paratenon is pathological, we are unaware of any evidence supporting this interpretation.
It is well known that knee joints undergo frequent movements in utero [6, 7, 11]. A baby usually has a knee flexion posture at birth. In new-borns and infants, stiffness at the knee is released from 21.4 to 3.3 degrees within the first 6 months [2]. This knee stiffness is apparently caused by a mass-effect of the thick paratenon armoured around the cruciate ligaments. A slight twisting or rotation at the knee is likely to release an infant’s knee because it can break the paratenon and accelerate cavitation. Finally, we had a study limitation that we did not perform a MOVAT’s pentachrome staining [17] for long-preserved specimens.
CONCLUSIONS
Differences in knee movements in utero seemed to cause differences in the thickness of the paratenon among foetuses. New-borns might have limited knee flexion due to a mass-effect of the thick paratenon around the cruciate ligaments. A slight twisting or rotation at the knee may help to release the knee, because it can break the foetal paratenon and accelerate cavitation.
