Abstract
The development of digital oedema, adhesion formation, and resistance to digital motion at days 0, 3, 5, 7, 9 and 14 after primary flexor tendon repairs using 102 long toes of 51 Leghorn chickens was studied. Oedema presented as tissue swelling from days 3 to 7, which peaked at day 3. After day 7, oedema was manifest as hardening of subcutaneous tissue. The degree of digital swelling correlated with the resistance to tendon motion between days 3 and 7. At day 9, granulation tissues were observed around the tendon and loose adhesions were observed at day 14. Resistance to digital motion increased significantly from day 0 to day 3, but did not increase between days 3 and 9. The early postoperative changes appear to have three stages: initial (days 0–3, increasing resistance with development of oedema), delayed (days 4–7, higher resistance with continuing oedema) and late (after day 7–9, hardening of subcutaneous tissue with development of adhesions).
Early mobilisation regimes involve active or passive flexion and extension of the repaired digit within the first several days after surgery (Coats et al., 2005; Kleinert et al., 1981; Kulkarni et al., 2006; May et al., 1992; Osada et al., 2006; Sirotakova and Elliot, 2004; Small et al., 1989; Tang, 2005). However, the factors contributing to the development of resistance to tendon motion are unclear. The optimum time for commencement of motion exercises after primary digital flexor tendon repair and optimal rehabilitation protocols remain controversial (Coats et al., 2005; Dowd et al., 2006; Elhassan et al., 2006; Osada et al., 2006; Sirotakova and Elliot, 2004; Tang, 2005). Ideally, postoperative tendon motion should counteract all forces resisting finger motion without rupturing the repaired tendon. Recently, the relationship between the resistance to tendon motion and digital oedema has been a topic of investigation (Cao and Tang, 2006). However, the intricate relations of changes in digital oedema, formation of postoperative adhesions and resistance to digital mobilisation in the postoperative period are still poorly understood.
The purpose of this study was to investigate the extent of postoperative digital oedema, the timing of adhesion formation and their association with resistance to digital mobilisation within 2 weeks after primary flexor tendon repair in a chicken model.
MATERIALS AND METHODS
One hundred and two long toes from both feet of 51 white Leghorn chickens weighing approximately 1.5 kg were used as the experimental model in this study. The flexor tendons in chicken toes are similar to those of human digits and are often used for investigation of digital flexor tendon surgery (Farkas et al., 1974; Tang et al., 2001; Xu and Tang, 2003). Bilateral long toes of chickens were used for the study. The study of the long toe flexors was divided into two parts. The first part of the study, involving 84 toes in 42 chickens, was used to observe digital oedema formation and tendon adhesion formation, and to examine histology of the oedematous volar soft tissues of the toes within 2 weeks after surgery. The second part, consisting of 18 toes in nine chickens, was used to score the digital oedema and to test the resistance to digital mobilization.
In the first part of the study, the toes were divided into non-operated normal control (eight toes), day zero control group (eight toes) and groups of 14 toes each which were evaluated at postoperative days 3, 5, 7 and 9, and a group of 12 toes evaluated at 14 days.
In the second part of the study, the 18 toes were divided into three groups of six toes each and evaluated at days 3, 5, and 7 (Fig 1).
Operative procedures
The chickens were anaesthetised by intramuscular injection with ketamine (50 mg/kg of body weight). The toes were operated on under sterile surgical conditions and tourniquet control using elastic bandages.
For the toes undergoing surgery, a zig-zag incision was made in the plantar skin of the bilateral long toes of the chickens between the proximal interphalangeal (PIP) and distal interphalangeal (DIP) joints, which is equivalent to flexor tendon Zone 2 in the human hand (Xu and Tang, 2003). The sheath was opened longitudinally by 1.0 cm to expose the flexor digitorum superficialis (FDS) and profundus (FDP) tendons. Two-thirds of both the FDS and FDP tendons were transversely lacerated from volar to dorsal with a fine scalpel blade. Partial tendon laceration was made in every toe by the same surgeon with the aid of an operating microscope. Consistency in the depth of laceration was addressed by putting a mark on the surface of the scalpel and making lacerations through the volar and middle parts of the tendons. Both the FDS and FDP tendons were repaired with 5-0 nylon sutures (Ethicon, Somerville, NJ, USA) by a modified Kessler method (Fig 2). Circumferential repairs were not added because the tendons were only partially cut and approximation of the cut ends was smooth after inserting the core suture. The sheath was laid back but not repaired, and the skin was closed with interrupted sutures. The long toe was immobilised by dressings and adhesive tapes in a semi-flexed position after surgery. The immobilisation mostly remained secure and the bandages were only torn away from the toes in two toes in two chickens during the postoperative immobilisation period. Thus, two additional chickens were used with one long toe from each being treated as previously described.
The immobilisation of the chicken toes in groups was released once a day in the last 2 days before the day of evaluation for those toes subjected to evaluation at different postoperative time points. After releasing the fixation, the operated toe was passively moved through a full flexion range 30 times and then placed back in the fixation. This was repeated on 2 consecutive days. On the third day, the chicken was killed and evaluated.
The toes of the chickens in the day 0 control group underwent the same surgery as others, but the toes were evaluated immediately following surgery. The toes in the normal control did not undergo surgery.
Evaluation of digital oedema
General changes of the oedema
In the first part of the study, we observed oedema formation of the toes from days 3 to 14, and recorded the temporal changes and appearance of the oedema in the volar soft tissues of the toes. The dorsal aspect of the toes is covered by squama, which is different from human hands. Thus oedema observed in this study was limited to the ventral aspect of the toes, and that on the dorsal aspect could not be assessed. Immediately after observation of digital oedema, the toes were subjected to biomechanical tests for resistance to digital motion.
Scoring severity of the oedema
In the second part of the study, oedema was only scored in the period from postoperative days 3, 5 and 7 when oedema took the form of “swelling” in this period. The digital oedema of six toes was scored on each of 3 days. The oedema was rated according to the grading criteria detailed in Table 1. The severity and locations of tissue oedema were graded separately for each of the operated digits (Cao and Tang, 2006), and given one of the following grades: (1) none, slight, moderate or severe oedema with respect to severity; and (2) none, limited, extended or extensive oedema with respect to location. An incremental point score of tissue oedema was given to each toe, combining scores of severity and those of extension. Oedema of all the toes was scored by the same surgeons. Immediately after scoring of the digital oedema, the gliding force of the FDP tendon and work of digital flexion were evaluated in the same way as in the first part of the study.
Biomechanical test of resistance to digital motion
Using toes from the experimental groups, evaluated from days 0 to 14, and the non-surgical control group, the end-point gliding force of the repaired FDP tendon and work of flexion of the operated toes were evaluated immediately after killing the animals. The toes were harvested by disarticulation through the knee joint level and, then, secured, with the toe tip pointing down, to a mounting board attached to the lower clamp of a tensile testing machine (Model 4411; Instron Corporation, Canton, MA, USA). Two K-wires were inserted through the metatarsal bone to keep the metatarsophalangeal joint extended and fixed, while the PIP, DIP and distal DIP joints were left free. The FDP tendon was exposed by an incision from the knee joint to the ankle joint level, and identified by its effect in flexing the distal DIP joint with a light pull. The FDS tendon was left alone without traction during the test. The proximal end of the FDP tendon was connected to the upper clamp of the overhead crossbar which was connected to a force transducer. A 0.5 N counterweight was attached to the long toe tip to ensure the digit fully extended and to take up the slack of the FDP tendon prior to starting the test.
After a preload of 0.1 N had been applied to the FDP tendon by the testing machine, the digit was flexed by advancing the overhead crossbar upwards at a constant speed of 25 mm/min until the FDP tendon excursion reached 18 mm. A fixed FDP excursion of 18 mm was used to assess the end-point force and work of toe flexion, because, with an FDP tendon excursion of 18 to 22 mm, the toes could be flexed over 80% to 90% of the total flexion arc (Cao and Tang, 2005; Xu and Tang, 2003). During tendon pulling, the load and displacement of the FDP tendons were measured simultaneously with a testing software (Series IX software; Instron Corp.). The load–displacement curve was displayed on a computer screen and the gliding force of the FDP tendon at the displacement of 18 mm was recorded. The energy was obtained by calculating the area under the curve, which represents the total work of forces that resist digit flexion (Peterson et al., 1986, 1990; Xu and Tang, 2003). The force, or work, of the first run on individual toes was recorded and used for analysis.
Observation of adhesion formation
After the above two evaluations, 2 cm soft tissues ventral to the digital flexor sheath centred by skin incision were taken from each toe for later histological observation detailed in the following paragraph. Then the volar flexor sheath was removed and the repaired FDP tendon exposed. The presence and status of granulation tissues, filmy adhesions or restrictive adhesions over the surface of the repaired FDP tendon was observed. Adhesions were recorded as “restrictive” when connections to the tendon surface at, or adjacent to, the repair site were observed to restrict passive motion between the tendon and the digital subcutaneous tissues or tendon sheath.
Histological observation of oedematous tissues
The volar soft tissues taken during observation of adhesions were placed in 10% formalin and underwent standard procedures of paraffin embedment. The tissues were sectioned longitudinally at a thickness of 4 μm and were stained with haematoxylin and eosin. Swelling of the tissues and deposition of fibrin in tissue spaces, as evidenced by increases in the diameter of collagen fibres and widening of spaces between these fibres, were observed under a microscope (Leica DMR 3000; Leica Microsystem, Bensheim, Germany).
Data analysis
The data of tendon gliding force and work of toe flexion were analysed statistically by a two-way analysis of variance. When the analysis indicated significance in the difference among data, a two-tailed Tukey test was used as a post hoc test. We set the level of significance at 0.05 and desired a statistical power of, or greater than, 0.80. In assessing the degree of association between oedema and the force or work, Pearson’s coefficient of correlation was calculated.
RESULTS
Changes of digital oedema in the first 2 postoperative weeks
From days 3 to 7, digital oedema presented chiefly as swelling of the volar skin and subcutaneous tissues. The skin and subcutaneous tissues gradually hardened after day 7. From day 9, though the diameter of the toes appears only slightly greater than that of the normal toes, the appearance of oedema in the toes was hardening of volar soft tissues. Thus the character of the oedema gradually changed from a diffuse compressible swelling in earlier days to a hardening of the tissues by day 9.
Digital oedema appears maximal on day 3, according to our scoring of the oedema. The mean oedema scores decreased progressively during the later days. Evaluated at postoperative days 3, 5 and 7, the oedema scores of the digits being passively moved for 2 days were 3.5±0.5, 2.8±0.7 and 2.6±0.7, respectively (Fig 3).
Histological changes in subcutaneous tissues
Compared with normal toes, the diameter of collagen fibres within the dermis was larger, and spaces between collagen fibres were greater at day 3. The widened fibres spaces probably indicate an increased amount of tissue fluid or fibrin deposition within the tissue. The increases in both diameter of the collagen fibres and the space between these fibres persisted in the sections of day 3 to day 14. From day 9 to day 14, the histological features of tissue oedema persisted when the digital swelling had substantially subsided and macroscopic digital oedema took the form of tissue hardening (Fig 4). The subcutaneous tissue also was thicker on days 3, 7 and 14 than in the normal toes.
Force of resistance to FDP tendon motion and work of digital flexion
The gliding force of the FDP tendons evaluated at days 3, 5, 7 and 9 increased significantly compared with that of the FDP tendons tested at day 0 (p<0.01 or p<0.001) (Fig 3). However, no statistical differences were found in the gliding force between the four postoperative time points. During testing, tendon rupture occurred in one toe of the day 3 group.
Similarly, we recorded significant increases in the work of flexion of the toes at days 3, 5, 7 and 9 compared with that of the toes on day 0 (p<0.05 or p<0.01). No statistical significance was found in the work between toes from 3 to 9 days after surgery.
The statistical powers were from 0.89 to 0.99, except that the power of comparison between the forces of tendon gliding evaluated at days 3 and 5 was 0.78.
Correlation of oedema with resistance to tendon motion
Oedema scores in individual toes correlated with the force and work in the toes moved at each of these days in varying degrees at days 3, 5 and 7. The correlation indices of the force and oedema scores were 0.611, 0.945 and 0.519 for the force of the toes evaluated at days 3, 5 and 7, respectively (Fig 5). Similarly, the correlation indices of the work and oedema scores were 0.442, 0.933 and 0.497 for the force of the toes evaluated at days 3, 5 and 7, respectively.
Observation of adhesion formations
No adhesion formation was identified in the samples from days 3, 5 and 7. Granulation tissue or filmy adhesions were seen in the tendons at day 9, but no well-formed adhesions around the tendons were detected. At day 14, loose adhesions were found around the repaired tendons, which, to some degrees, were restrictive to tendon gliding.
DISCUSSION
Oedematous change after surgery is almost inevitable and has long been considered to be a factor contributing to resistance to gliding of the surgically repaired tendon (Elliot et al., 2007; Kleinert et al., 1981; Tang, 2007), but the specific role of oedema with respect to resistance to motion of surgically repaired tendons has not been clear. In a chicken tendon injury and repair model, oedema was recorded at multiple time points after surgery. Due to species differences, oedema noted in the chicken toes was on the volar side of the toes, rather than circumferentially around the digits, because the dorsal aspect of the toes is covered by squama.
Two macroscopic forms of oedema were observed in the digits after surgery, viz., tissue swelling and tissue hardening. Swelling of the tissues developed earlier after injury, approximately from 3 to 7 days post surgery, which was followed by hardening of the tissues from day 9 to the end of our observation period (day 14). Our scoring of the oedema was limited to the first phase when oedema presented chiefly as tissue swelling, because this scoring system was developed to record degrees of tissue swelling (Cao and Tang, 2005), thereby not applicable to oedema in the form of tissue hardening. Oedema in the form of tissue hardening was quite consistent, which did not appear to be much different among the toes. In this experiment, oedema in the form of tissue swelling peaks at day 3, subsides at day 7 and almost disappears at day 9. After that, tissue hardening develops, with a mild increase in diameter of the digits.
From day 3 to day 9, we did not find any significant differences in the tendon gliding force and work of digital flexion. This appears reasonable. This implies that the resistance caused by tissue swelling in the earlier days and the resistance caused by tissue hardening and extensor tethering (both relate to increased fibrin deposition in this period) are basically equal. It should also be noted that the oedematous responses likely differ between the tendon and its volar soft tissues; decreases in severity of the digital swelling after the first several days after surgery only signal a decline in acute biological responses of the skin and subcutaneous tissue, not necessarily a decline in the responses or swelling of the tendons. The biological responses and swelling of the repaired tendons may increase when swelling of subcutaneous tissue subsides. When swelling of volar soft tissues decreases, tethering of the extensor mechanism probably contributes increasingly to the resistance to active tendon motion as days of post-surgical digital immobilisation pass by (Kulkarni et al., 2006). At day 9, granulation tissues or filmy adhesions were observed around the tendons – these did not cause observable resistance to tendon motion.
A significant increase in the resistance to digital motion at day 14, when loose adhesions formed around the tendons, was recorded. The increase in the resistance could be due to a combination of the effects of formation of adhesions, increased extensor tethering and stiffness of finger joints, as the length of the postoperative immobilisation period increases.
The possible application of this study clinically is to inform potential flexor tendon rehabilitation protocols. Fig 6 summarises the findings from this study, including speculative changes in other factors, such as extensor tethering or stiffness of finger joints, and our intended divisions of the early postoperative period. There could be potential merit in resting the repair in the first few days and then starting mobilisation in what we refer to as the delayed period. This would not adversely affect motion from the point of view of adhesion formation.
Though movement of the repaired digits to disrupt adhesion formation appears unnecessary before day 9, extensor tethering becomes more obvious as the number of days of immobilisation increases. Therefore, we feel that ideally postoperative motion can start from postoperative day 4 or 5. It is probably also acceptable to start digital motion as late as on day 7 or 9 for the patients who are unable to start digital motion earlier.
Past reports have identified that the majority of ruptures of the primarily repaired tendons occur in the first 2 weeks after surgery (Baktir et al., 1996; Dowd et al., 2006; Elliot et al., 1994; Small et al., 1989). The ruptures occurred most easily in a period overlapping softening of the tendon ends, which is approximately in the second week after surgery. Small et al. (1989) reported two ruptures in the first week, five in the second week, one in the third week and three in the sixth week out of 138 Zone 2 flexor tendon repairs. In the report of Elliot et al. (1994), among 13 fingers and five thumbs with ruptures of the repairs, five occurred in the first week, seven in the second week, three in the third week, two in the fourth week and one in the fifth week. Dowd et al. (2006) recorded six ruptures in the first week and 23 ruptures in the second week out of 62 ruptures within 9 weeks after primary repairs. Delaying digital motion exercise for several days cannot be expected to totally eliminate the incidence of rupture.
In this study, we scored oedema when it presented chiefly in the form of digital swelling from days 3 to 7. On each day, the severity of the oedema in the digits roughly correlated with the resistance to digital mobilisation. Higher scores of oedema indicate an increase in the resistance to tendon motion. Therefore, an oedema scoring system similar to that used in this study probably can help surgeons to judge the severity of digital oedema clinically. In the light of findings in this study, the resistance to tendon motion in the operated digits varies among the digits and should be judged individually. Motion exercises may need to be applied cautiously, and to limit the range, frequency and speed of active motion of the fingers in the presence of severe oedema.
A chicken model was used in the present study. Though this is a proven model for flexor tendon experimentation and resembles Zone 2 flexor tendon anatomy in human beings more closely compared to other animal models (Su et al., 2006; Wong et al., 2006), we created only a two-third cut in the flexor tendons to avoid rupture of the repaired tendon during simulated postoperative motion of the digits. Another reason for the use of a partial, rather than complete, tendon cut model was to ensure sufficient tendon strength for performance of the tests. Test of work of flexion of chicken toes usually needs about 10 N traction, but the healing strength of a completely cut flexor tendon was much smaller (about 5 N) at the first 2 post-surgical weeks in our pilot test. We have used the partial tendon laceration previously to assess tendon adhesions and it has proved satisfactory (Tang et al., 2001; Xu and Tang, 2003). Nevertheless, the healing responses of the tendon could be less than those in a completely lacerated tendon.
One of the limitations of our study is that we assessed the resistance or oedema at six selected time points, viz. postoperative days 0, 3, 5, 7, 9 and 14. It would have been more informative to evaluate the results on each postoperative day. Secondly, the current study was one of observation of oedema of the skin and subcutaneous tissue, rather than oedema of the tendon. Assessment of the oedematous status of the tendon is impossible clinically after flexor tendon repairs, but this appears to be a good topic for a future study aimed at characterising the oedema of the repaired tendons and its relation to tendon gliding resistance.
Finally, factors pertaining to the model may have created limitations to the study. The study used an animal model, and the time course of healing responses and oedematous change in animal digits may be different from those of human hands. A technical difficulty in designing this study was whether we should standardise the period of simulated tendon motion or the period of toe fixation prior to motion in this model, which necessitates evaluation at different post-surgical days. We chose to standardise the period of tendon motion by passive digital motion for two consecutive days before evaluation. Gliding resistance and oedema formation could be different from our present findings if the toes were immobilised for a fixed period followed by motion for different days. Also, the oedema could only be judged subjectively with respect to scoring criteria. These points should be taken into consideration when interpreting current findings.
Footnotes
Figures and Table
Supported by grants from Health Bureau of Jiangsu Province, Fund to Jiangsu Hand Surgery Center, and Department Fund of the Affiliated Hospital of Nantong University.