ELSEVIER Journal of Orthopaedic Research 22 (2004) 990-997

Journal of Orthopaedic

Research www.elsevier.com/locate/orthres

Tendon injury response: assessment of biomechanical properties, tissue morphology and viability following

flexor digitorum profundus tendon transection Matthew J. Silva *, Timothy M. Ritty, Konstantinos Ditsios, Meghan E. Burns, Martin I. Boyer, Richard H. Gelberman

Orrhopuedic Research Laboratories, Depurtment of’ Orthopuedir Surgery. Barnes-Jewish Hospitat at Washington University, I Barnes-Jewish PIazu, Suite 1100 WP, St. Louis, M O 63110, USA

Abstract

Insertion site injuries of the flexor digitorum profundus (FDP) tendon often present for delayed treatment. Apart from gross observations made at the time of surgery, the changes that occur in the flexor tendon stump during the interval from injury to repair are unknown. These changes may include tendon softening and loss of viability, which may contribute to the poor outcomes ob- served clinically and experimentally. Thirty-eight FDP tendons from 23 adult dogs were transected sharply from their insertions on the distal phalanges and were not repaired. Dogs were allowed full weight bearing and were euthanized 7 or 21 days after injury. Biomechanical testing indicated that the resistance of injured tendons to pullout of a Kessler-type suture was not different from control tendons at I days and was increased at 21 days by 25% (p < 0.05). Morphologically, at 7 and 21 days the cut surface had a smooth appearance and the end of the injured tendon was increased in thickness by 30% and 50%, respectively (p < 0.05). His- tologically, we observed increased cellularity and dramatic fibroblast proliferation within the injured tendon stump; there was no evidence of decreased cell viability. We conclude that during the interval from 0 to 21 days after FDP insertion-site injury, tendons cells are viable, proliferative and synthesizing new matrix. This leads to increased tendon size and enhanced resistance to suture pullout. These findings offer a scientific rationale supporting the clinical practice of surgical re-attachment within the first 3 weeks after injury. 0 2004 Orthopaedic Research Society. Published by Elsevier Ltd. All rights reserved.

Introduction

Tendons that are injured at their insertions are often not immediately repaired [2,6,13,17]. In particular, avulsions of the flexor digitorum profundus (FDP) tendon often go unrecognized after injury, thereby necessitating delayed treatment. If the distal tendon stump does not retract into the palm, the treatment of choice is surgical re-attachment of tendon to bone [9,10,13,14,16]. There are few data on clinical outcomes following delayed repair of FDP insertion site injury, although past reports indicate that loss of motion and grip strength may occur [13,14] and in our clinical experience a subset of patients do not recover full function. Results from an experimental study of delayed repair following FDP tendon midsubstance injury sup-

* Corresponding author. Tel.: +1-314-362-8597; fax: +I-314-362-

E-mail address: silvam@msnotes.wustl.edu (M.J. Silva). 0334.

port these clinical observations [7]. Three weeks after repair, canine FDP tendons that were repaired 7- 21 days after injury had significantly reduced range of motion compared to tendons repaired immediately. Tendons repaired 21 days after injury also had a 40% decrease in ultimate force (although this difference did not reach statistical significance). Because tendons were not examined at the time of repair, it is not known whether or not changes occurred in the tendon tissue during the delay interval that may have contributed to the negative outcomes.

Except for gross observation made at the time of surgery, the changes that occur in the flexor tendon stump during the interval from injury to repair are un- known. Enlargement of the tendon stump has been noted [5 ] , but it is not clear if this is due to edema or to new matrix synthesis. Grossly, the tendon stump may appear viable [14], although the viability of internal fi- broblasts has not been demonstrated. Of particular rel- evance to insertion site injuries is that the blood supply

0736-0266/$ - see front matter 0 2004 Orthopaedic Research Society. Published by Elsevier Ltd. All rights reserved. doi:lO. 101 6/j.orthres.2004.01.004

M.J. Silau el al. / Journul of Orihopuedic Rrseurch 22 (2004) 990-YY7 99 1

of the normally vascular distal tendon [15] is disrupted and the extrasynovial portion of the tendon is translo- cated to the intrasynovial sheath. These changes, along with the loss of normal tensile forces, may lead to ten- don softening prior to repair. However, no biome- chanical data have been reported that address this important clinical issue.

Our objectives were to describe the changes in tendon substance that occur following a complete laceration of the FDP tendon from its insertion, and to determine whether or not there are alterations in tendon strength or viability that might negatively impact its repair potential. We transected canine FDP tendons from their insertions and left them unrepaired. At 7 and 21 days after injury, we assessed the mechanical competence of the tendons using suture pullout tests and evaluated cell viability and tendon morphology using light and elec- tron microscopy.

Materials and methods

Studies were performed using 38 injured and 38 control tendons from 23 adult mongrel dogs (23-27 kg body mass: Covance Research Products, Alice, TX) and were approved by our institutional animal studies committee. For surgery, each dog was anesthetized using thiopental sodium (0.5 ml/kg IV), atropine (0.5 ml) and acepromazine (0.2 ml), then incubated and maintained on 1% isoflurane. The right forelimb was shaved, washed with Betadine and exsanguinated: sur- gery was performed under tourniquet control. Midlateral incisions were made on the second and fifth digits of the right forelimb to expose the insertions of the flexor digitorum profundus (FDP) tendons into the volar base at the proximal aspect of the distal phalanges. The tendons were released directly from their bony insertion by sharp dissection and the vinculum brevis profundus was cut, simulating a clean laceration with complete interruption of the blood supply to the distal tendon. Tendons were not repaired and the wound was closed using nylon suture. Following surgery, dogs were allowed unrestricted cage activities and appeared to be ambulating normally on the first post-operative day. Eleven dogs were killed 7 days and 12 dogs were killed 21 days after surgery by overdose of pentobarbital. These times matched the delay intervals reported in a previous study [7]. The right (injured) and left (control) forelimbs were amputated at the elbow joint.

Prior to surgery, injured tendons had been assigned to one of four outcome groups. At 7- and 21-day time points, 10 tendons were des- ignated for biomechanical testing, 3 tendons for viability analysis, 3-5 tendons for histological analysis, and 2 tendons for scanning electron microscopy. An equal number of contralateral tendons were used for controls. (Eight injured tendons from 8 dogs were used for another study not reported here.) At necropsy, the injury site was inspected: a single observer subjectively graded the prevalence of adhesions be- tween tendon and sheath as none, mild (one or two immature adhe- sions) or moderate (three or more immature adhesions, or one or more dense adhesions).

Biomechunical testing

Tendons were frozen in situ a t -20 "C until testing. After the forelimbs were thawed to room temperature (21-23 "C), the FDP tendons from the second and fifth digits were cut at the level of the musculotendinous junction and isolated. We grasped the distal tendon end (the injury site), using a modified Kessler stitch with 4-0 looped suture (Supramid Extra, S. Jackson, Inc., Alexandria, VA): the grasping loops were placed 1 cm from the tendon end. The free ends of the suture (two double-stranded limbs) were tied to form a loop approximately 3 cm in length. A modified Kessler stitch was chosen

based on its relevance to clinical suture methods. Contralateral control tendons were prepared in a similar fashion. Tendon dimensions (medial-lateral width; dorsal-volar thickness) were measured at the distal end using digital calipers (0.01 nini resolution: 0.2 mni accuracy).

Tendon-suture pullout tests were performed using a servohydraulic materials testing machine (Instron 8500R, Canton, MA) and a custom gripping apparatus. The proximal tendon end was gripped in a soft tissue clamp at a distance of 4 cm from the distal end. The suture loop on the distal end was looped once around a steel rod, which was passed through two brass bushings attached to a yoke-shaped fixture that was attached to the testing machine. The rod was free to rotate within the bushings (like a low-friction pulley), allowing for equal force to de- velop within the two arms of the suture loop as the tendon-suture construct was tensioned. A tensile pre-load of I N was applied and the specimen was then elongated at 0.5 mmls until failure. From force- elongation plots we determined: maximum force, displacement at maximum force, displacement-to-failure and energy-to-failure. Re- peated measures analysis of variance testing (Statview 5.0, SAS Insti- tute, Cary, NC) was used to determine the effects of injury (injury vs. control) and time (7, 21 days) on biomechanical properties and tendon dimensions. If significance was noted (p < 0.05), post-hoc comparisons (Tukey tests) were made between injured and control groups at each time point.

Scunning electron microscopjl

For scanning electron microscopy, the distal 2-2.3 cm of the F D P tendons were collected immediately postmortem and fixed with 2.5% glutaraldehyde in isotonic buffer for 24 h at 4 "C, followed by post- fixation with 1% OsOd in isotonic buffer for 2 h. The samples were then rinsed, dehydrated in ascending concentrations of ethyl alcohol, and critical point dried from liquid CO2. Mounted samples were coated with 400-500 A of gold, and examined in a Hitachi S-450 scanning electron microscope operated at 20 kV.

Viability stuining

Cell viability was assessed using a live/dead staining protocol modified from one reported previously [I]. The distal FDP tendon stump was removed from anesthetized animals prior to euthanasia (to minimize cell death prior to labeling) and from each tendon a full- thickness midsagittal section 1 mm thickness by 2 cm length was cut by hand using a razor. Specimens were incubated at 37 "C in Dulbecco's minimal essential medium (DMEM) containing 1 pM ethidium ho- modimer for 5 min, washed in DMEM and then incubated in a solu- tion of 1 pM cakein-AM for 45 min and washed again. Reagents were obtained from a commercial kit (LIVE/DEAD Kit L-3224, Molecular Probes, Inc., Eugene, OR). Specimens were placed in DMEM between glass slides and examined immediately under a fluorescent microscope (Zeiss AxioPhot, Carl Zeiss. Inc., Thornwood, NY). Representative images were captured with a digital camera (Zeiss). With this protocol, viable cells are stained with calcein-AM and appear green when excited by blue light (480 nm), while dead cells are stained with ethidium homodimer and appear red when excited by green light (535 nm). Specimens were qualitatively evaluated by two examiners.

General hDtologj1

The distal 2-3 cm of the F D P tendon stump was dissected within 1 h postmortem and fixed in buffered lo'%) formalin at 4 "C for 24 h. Samples underwent stepwise dehydration, were embedded in paraffin and sections were cut at 5 pm thickness. Routine hematoxylin and eosin staining was performed. Sections were examined by light microscopy and images were captured with a digital camera.

Results

At the time of necropsy, the distal end of the FDP tendon from each injured digit had retracted into the synovial sheath. At 7 days, the distal ends of 16 of 18 tendons were located between the proximal and distal

992 M.J. Silm ef ul. I Journal of Orthopurdrc Reseurch 22 (2004) 990-997

pulleys, while two were located beneath the distal pulley. (Note that the proximal and distal pulleys in the canine correspond to the human A2 and A4 pulleys, respec- tively.) At 21 days, the distal ends of 12 of 20 tendons were located between the proximal and distal pulleys, while six were located beneath the distal pulley and two were located beneath the proximal pulley. Adhesions between the FDP tendon and sheath were rare at 7 days (16 of 18 tendons had no adhesions; 2 had mild adhe- sions; 0 had moderate adhesions) but common at 21 days (3 of 20 tendons had no adhesions; 15 had mild adhesions; 2 had moderate adhesions). We saw no evi- dence of disruption of the proximal mesotenon.

In those tendons in which the end was located be- tween the pulleys, the distal 5 mm appeared enlarged, while in those tendons in which the end was trapped under the pulley, the end did not appear grossly en- larged. Despite this variability, caliper measurements indicated that the distal tendon end was, on average, significantly enlarged in injured tendons. Injured ten- dons were increased in thickness by 30% at 7 days 0, = 0.013) and 50%) at 21 days (p < 0.001) compared to contralateral controls (Table 1). In addition, there was a trend for increased tendon width at 21 days (p = 0.098).

Tensile testing indicated that the resistance to suture pullout was not diminished in injured tendons compared to controls (Table 2). Rather, the maximum force in injured tendons at 21 days was increased by 25%) com- pared to control (p = 0.016). Maximum force at 7 days was not different between injured and control tendons (JJ = 0.88). Displacement at maximum force did not differ significantly between injured and control tendons at either 7 or 21 days (p = 0.75). The only parameter that showed a negative change with injury was the dis- placement-to-failure (overall p = 0.038), which was de-

creased by 140/0 at 21 days (p=O.O64). However, the decreased displacement-to-failure did not result in a significant decrease in energy-to-failure between injured and control tendons (p = 0.52). The failure mode was predominantly one of suture pullout from the tendon, with the suture loop intact; 38 of 40 specimens failed by this mode. Two specimens (both in the 21-day, injury group) failed by rupture of the suture loop away from the knot (maximum force 48.8 and 45.7 N).

Scanning electron microscopy revealed that injured tendons were enlarged at the distal end, with a smooth surface. In control specimens that were lacerated post- mortem, the cut end of the tendon and the site of ten- don-bone insertion could be easily distinguished from the adjacent, uninjured tendon surfaces (Fig. 1). High power ( 2 0 0 0 ~ ) views of control tendons revealed thick collagen fibers that were exposed by the surgical tran- section. In contrast, 7 and 21 days after injury the cut end of the tendon had a smooth appearance and it was difficult to identify the margins of the original tendon- bone insertion. High power views showed the presence of new collagen fibrils and other unidentified matrix proteins on the tendon surface covering the distal end. The morphological differences between the control and injured tendons appeared to be due to the proliferation of cells and the deposition of new extracellular matrix proteins over the former tendon-bone interface, extend- ing several mm proximally in the injured specimens. Examination of the tendon 1-2 cm proximal to the laceration site revealed no differences between control and injured tendons.

Viability staining revealed large number of living cells at the injury site and no evidence of increased cell death in injured tendons. Seven and 21 days after injury, ten- dons displayed intense fluorescence at the distal end of

Table 1 Dimensional data from measurement of the distal tendon end ( n = 9-10 per group)

Parameter 7-Day 2 1 -Day

Control Injury Control Injury Dorsal-volar thickness (mm) 1.7 f 0.3 2.2 f 0.6'* 1 . 8 f 0 . 3 2.7 f 0.6** Medial-lateral width (mm) 4.1 f 0 . 3 4.2 f 0.5 4.3 f 0.6 4.8 f 0.3'

* p < 0.10; **p < 0.05 vs. contralateral control.

Table 2 Biomechanical data from suture pullout tests ( n = 9-10 per group)

Parameter 7-Day 2 1 -Day

Control injury Control Injury Maximum force ( N ) 30.6 f 6.6 31.1 f 6 . 8 30.3 t 5.0 38.0 f 7.8** Displacement to maximum force (mm) 14.2 f 3.6 10.6 f 5.6 11.4 f 5.7 14.2 24.1 Displacement-to-failure (mm) 21.3 f 3.0 20.0 f 2.8 21.6? 2.4 18.6f4.1' Energy-to-failure (N mm) 4 l S f I06 370 f 102 381 f 7 4 391 f 115

* p < 0.10; =*p < 0.05 vs. contralateral control.

M. J. Silw rt ul. I Journul of Orthopurdic Rr.vearcli 22 (2004) 990-997 993

Fig. 1. SEM photomicrographs illustrating surface morphology of control and injured tendons. The control tendon was lacerated post- mortem, and its insertion site (white oval) exhibited a frayed appear- ance, with exposed collagen bundles observed at high power. In contrast, tendons injured 7 or 21 days prior to death had enlarged ends and smooth surfaces. Newly synthesized matrix appeared to be cov- ering the distal tendon surface. (Boxes in 20x images indicate the regions depicted in high power panels.)

the tendon stump due to the large number of living cells at this site (Fig. 2a). The viability did not appear to be influenced by the presence or absence of adhesions. For example, at 21 days no loss of viability was seen in the one specimen (one of three) that did not have adhesions. The greatest fluorescence was confined to the distal 1 cm of the tendon stump and was much greater in injured tendons than in controls. High power views revealed that cells at the injured end had a spindle shaped fibroblastic appearance and were oriented randomly, in contrast to the longitudinal orientation of cells in the control tendon. One to two centimeters proximally lin- ear arrays of living tendon cells were present in both control and injured tendon samples. Within injured tendons, however, the cell arrays could often be seen in a wavy alignment, rather than in a linear longitudinal

alignment (Fig. 2b). We saw no evidence of decreased cell viability in injured tendons. Qualitatively, the num- ber of dead cells were similar in both control and injured tendons, likely due to postmortem manipulation rather than to in vivo necrosis (Fig. 2b). Small blood vessels in some control tendons were illuminated by calcein-AM stain (Fig. 2c). Blood vessels were not seen in any ten- dons at 7 days but were again apparent at 21 days, both within adhesions and in the distal aspect of the tendon stump. Adipocytes were sometimes present at the ten- don end at both 7 and 21 days.

Hematoxylin and eosin staining of the distal aspect of the tendon stump 7 and 21 days after injury demon- strated no evidence of necrosis or lysed red blood cells. To the contrary, vigorous cell proliferation and scar formation were evident, although the degree of the in- jury response was variable at both time points. A con- sistent observation was that a “cap” of proliferative tissue enveloped the tendon end (Fig. 3a and b), and that there was markedly increased cellularity in the cap compared to the tendon body (Fig. 4). Within the distal 4 cm of the 7- and 21-day injured tendons, the collagen and the cell arrays were wavy as compared to the more linear appearance of the controls. Near the site of injury, proliferation of the visceral sheath cells appeared to contribute to the outer cell layers of the scar area (Fig. 3a). Adipocytes were present at the outermost cell layer at 7 and 21 days. The cells within the scar appeared to arise from both the outer cell layers as well as the en- dotenon arrays. Examination of tendons that had formed adhesions 21 days after injury revealed evidence of angiogenic activity. Within scars areas, there was abundant fibrin along with extravasated red blood cells (Fig. 3b). Of note, no differences were observed between the 21-day specimens that had no adhesions (two of five) and those that had adhesions (three of five).

Discussion

Several factors may negatively influence healing fol- lowing delayed FDP insertion-site repair, including loss of tendon tissue quality, loss of bone mass [3], and re- pair-site gap formation [ 181. In addition, shortening of the muscle-tendon unit may lead to increased tendon tension after re-attachment, altering digital function and putting the repair-site at increased risk for gapping. Of these factors, we focused on tendon tissue quality in this study. We evaluated the changes in tendon biomechan- ical properties, morphology and cell viability following a complete FDP tendon laceration at the insertion site. The resistance to pullout of a clinically relevant Kessler- type suture was not diminished in the injured tendon compared to control, with the maximum pullout force increased by 25% at 21 days. The distal end of the in- jured tendon was increased in size at 7 and 21 days and

994 M. J. Siloa et al. I Journal of Orthopaedic Research 22 (2004) 990-997

Fig. 2. Representative fluorescent photomicrographs ( lox, except as noted) of interior midsagittal sections from whole tendons stained with calcein- AM (green) or ethidium homodimer (red) to detect living and dead cells, respectively. (a) The distal ends of 7- and 21- (shown) day tendons revealed large, dense groups of viable cells (TOP) with little evidence of necrosis (MIDDLE). Fibroblast-like cells that appeared to be proliferating were consistently observed at the end of the injured tendons (BOTTOM, 20x). (b) One to two centimeters proximal to the laceration site, cells within control tendons displayed a characteristic linear array appearance, while the cell arrays in injured tendons took on a wavy configuration. Quali- tatively, the numbers of dead cells appeared similar in control and injured tendons. (c) Microvasculature was observed in some control tendons, but not samples evaluated 7 days after injury. Newly formed blood vessels and adipocytes (arrows) were observed in adhesion tissue after 21 days.

the cut surface had a rounded-off, smooth appearance. The distal tendon end showed evidence of increased cellularity and dramatic fibroblast proliferation, pri- marily due to expansion at the epitenon surface. There was no evidence of decreased cell viability over the distal 2-3 cm region that we examined, either on the surface or in the interior of the tendon. Taken together, our results indicate that there were no changes in the tendon sub- stance at the injury site during the first 21 days that would impair the potential for a successful delayed repair.

There have been few reports describing the changes that occur in tendon tissue following insertion site in- jury, and these have been limited to clinical observations made at the time of surgery. Leddy and Packer [14] and Leddy [ 12,131 classified injuries to the FDP tendon-bone insertion site into three types based on the level of retrac- tion and bony involvement. Most common are those in which the tendon stump remains within the digital sheath (type 11), while less common are those in which the tendon retracts into the palm (type I) and those associated with fracture of the distal phalanx (type 111).

Control

M . J. Si1t.u el ul. I Journal of Orthopaedic Research 22 (2004) 990-997

7 day

995

21 day

Fig. 3. Photomicrographs of tendon sections stained with hematoxylin and eosin. (a) Control and injured tendons after 7 and 21 days (4x). After transection, varying amounts of visceral tissue remained. After 7 and 21 days, evidence of a vigorous injury response was seen; there was no evidence of lysed erythrocytes or necrosis. Adipocytes were sometimes present at the outer layers of the injured end (arrows). (b) Tendon end 21 days after injury. Image is representative of tendons that formed adhesions with angiogenic activity penetrating the tendon scar area. Low power image (4x) shows red blood cell extravasation in newly synthesized scar region (arrow). Adhesion tissue is not shown in this section. Higher power image O OX), taken from region designated by the yellow box, shows intact red blood cells (red ovals) and extensive fibrin clot formation (pink).

Based on the Leddy classification, our canine model most closely approximates the more common type I1 insertion site injury, in which the distal end of the ten- don stump retracts but remains within the sheath. In our model we have disrupted the distal blood supply by severing the vinculum brevis profundus, but have not interrupted the blood supply proximal to the sheath [4]. A similar level of vascular insult probably accompanies type I1 injury in patients. Using this clinically relevant animal model, we are able to describe, for the first time, the natural progression of tendon changes following an FDP insertion site injury that is not immediately re- paired.

Our results support the observations of Leddy and Packer [14], who noted that in type I1 injuries the tendon end appears to remain viable after injury. Moreover, our findings extend these clinical observations in several important ways. First, they demonstrate that both external and internal tendon fibroblasts remain viable up to 21 days. Second, they indicate that a normally vascularized, extrasynovial segment of tendon [ 151 can be detached from its blood supply, translocated to the intrasynovial sheath and not only survive, but show dramatic evidence of synthesis and proliferation up to 3

weeks after injury. Third, and perhaps most important, our results indicate that the tendon stump does not “soften” but has increased resistance to suture pullout 21 days after injury, consistent with increased tendon thickness. However, we do not know if the increased pullout strength that occurs in the interval from injury to surgical repair would be sufficient to offset the pos- sible increase in tendon tension caused by muscle-ten- don shortening.

A second clinical observation supported by our findings is that the distal tendon stump may become so enlarged that it is difficult to pass through the A4 pulley during re-attachment [5]. Our histological analyses indi- cate that this increase in size is due to matrix synthesis and cell proliferation rather than to edema as had been proposed [5].

The histological changes that we observed following FDP tendon insertion site injury are similar to changes reported in experimental studies of tendon inidsubstance injury and repair. Gelberman et al. [S] reported in- creased cellularity and matrix synthesis in the first 6 weeks after midsubstance FDP tendon injury and repair. They noted that in mobilized tendons, epitenon cells were primarily responsible for the healing response

996 M. J . Silvu et al. I Journal of Orlhopaedic Research 22 (2004) 990-997

Control

7 day

21 day

Fig. 4. A dense cellular cap enveloped the injured FDP tendon tip. Hoechst 33258 staining viewed with UV illumination revealed the relatively dense cellularity of the cap that forms over the tendon ends after injury ( l o x ; cell nuclei are blue).

although some proliferation of endotenon fibroblasts was also observed. Kakar et al. [ 1 I] also reported in- creased cellularity in the epitenon in the first week after partial laceration injury in rabbits, with increased cel- lularity of the endotenon observed at 12 weeks. A common observation of these previous studies [8,11] and our current study is that the epitenon rapidly prolifer- ates to encapsulate the exposed injury surface. The cel- lular contribution from the endotenon to the encapsulating cell layer appears to be minimal, although the endotenon appears to contribute significantly to

interior scar formation. Taken together, these compar- isons suggest that the FDP tendon response after insertion site injury does not differ notably from the response following injury and repair in the tendon midsubstance, at least during the first 21 days.

Several limitations should be noted when interpreting our results. First, our findings relate only to injured tendons that remain in the intrasynovial sheath (type 11). Because we did not simulate a type I injury, we cannot address the issue of possible necrosis of the FDP tendon following retraction into the palm. Second, the changes we observed following sharp laceration injury may not replicate the changes following an avulsion injury. Based on preliminary trials, we were unable to reproducibly create an avulsion model in the canine forelimb. A surgical laceration cannot recreate the trauma of an avulsion injury, but our method has the advantage of being reproducible and appropriate for use in a large animal model while recreating the loss of tendon load and blood supply that would result from avulsion. Whereas our study may have only indirect relevance to avulsions of the FDP tendon, it has direct relevance to lacerations at or near the insertion site (Zone I) that are not immediately repaired. Third, the tests we used to assess mechanical competence of the tendon stump do not allow us to assess tendon mechanical properties per se, but rather those of a tendon-suture construct. It is not possible to grip the end of the tendon in such a way as to assess tendon strength and stiffness in the region of interest, i.e., the distal 1 cm. Therefore, we chose to use a suture pullout tests that would allow us to assess the ability of the tendon to sustain a clinically relevant re- pair method.

One clinical implication of our study is that there is neither a histological nor a biomechanical basis for resecting the tissue at the distal tendon stump at the time of repair. Our findings did not reveal evidence of tendon necrosis in the first 3 weeks after insertion site injury. To the contrary, the tendon end was covered with viable, proliferating cells that appeared to be producing matrix. These findings suggest that, as much as surgical condi- tions allow, the tissue at the distal end of the tendon can be left intact at the time of repair as a rich source of synthesizing cells.

Finally, since we observed no deleterious changes in the tendon tissue, our results provide a scientific ratio- nale supporting the common clinical practice of surgical re-insertion within the first 3 weeks after injury [9,10, 13,14,16]. Moreover, they indicate that other factors must be responsible for the poor functional outcomes observed in some patients [13,14] and in experimental studies of delayed repair [7]. These factors were not ad- dressed in this study, but may include shortening of the muscle-tendon unit, repair-site gap formation [ 181, bone resorption at the insertion site [3], and loss of motion due to adhesion formation.

M. J. Silvu et ul. I Journul of Orthopaedie Research 22 (2004) 990-997 997

Acknowledgements

The authors acknowledge the contributions of Tim- othy Morris for animal care and Mike Vieth for electron microscopy. We thank Dr. Robert Mecham of the Department of Cell Biology and Physiology for use of microscopy facilities and Dr. Erika Crouch of the Department of Pathology and Immunology for review- ing the histological slides. Supported by grant AR33097 from the National Institute of Arthritis and Musculo- skeletal and Skin Diseases.

References

[I] Ark JW, Gelberman RH, Abrahamsson SO, et al. Cellular survival and proliferation in autogenous flexor tendon grafts. J Hand Surg-Am Vol 1994;19:249-58.

[2] Ballas MT, Tytko J, Mannarino F. Commonly missed orthopedic problems. Am Fam Physician 1998;57:267-74.

[3] Ditsios K, Boyer MI, Kusano N, et al. Bone loss following tendon laceration, repair and passive mobilization. J Orthop Res 2003; 21:990-6.

[4] Ditsios K, Leversedge FJ, Gelberman RH, et a]. Neovasculariza- tion of the flexor digitorum profundus tendon after avulsion injury: an in vivo canine study. J Hand Surg [Am] 2003;28:231-6.

[5] Elliot D, Khandwala AR, Ragoowansi R. The flexor digitorum profundus demi-tendon-a new technique for passage of the flexor profundus tendon through the A4 pulley. J Hand Surg-Brit Vol 2001;26:422-6.

[6] Ellman H, Hanker G, Bayer M. Repair of the rotator cuff. J Bone Joint Surg [Am] 1986;68:113&44.

[7] Gelberman RH, Siege1 DB, Woo LYS, et al. Healing of digital flexor tendons: Importance of the interval from injury to repair. J Bone Joint Surg 1991;73-A:66-75.

[8] Gelberman.RH, Vande Berg JS, Lundborg GN, Akeson WH. Flexor tendon healing and restoration of the gliding study. An ultrastructural study in dogs. J Bone Joint Surg [Am] 1983;65:70- 80.

[9] Gunter GS. Traumatic avulsion of the insertion of flexor digitorum profundus. Aust New Zealand J Surg 1960;30:2-8.

[lo] Honner R. The late management of the isolated lesion of the flexor digitorum profundus tendon. Hand 1975;7:1714.

[ I I ] Kakar S, Khan U, McGrouther DA. Differential cellular response within the rabbit tendon unit following tendon injury. J Hand Surg-Brit Vol 1998;23:627-32.

[I21 Leddy JP. Avulsions of the flexor digitorurn profundus. Hand Clin 1985; 1:77-83.

[I31 Leddy JP. Flexor tendons-acute injuries. In: Leddy Green DP, editor. Operative hand surgery, vol. 3. New York: Churchill Livingstone; 1993.

[I41 Leddy JP, Packer JW. Avulsion of the profundus tendon insertion in athletes. J Hand Surg 1977;2:6&9.

[I51 Leversedge FJ, Ditsios K, Goldfarb CA, et al. Vascular anatomy of the human flexor digitorum profundus tendon insertion. J Hand Surg [Am] 2002;27:806-12.

[I61 Reef TC. Avulsion of the flexor digitorum profundus: an athletic injury. Am J Sports Med 1977;5:281-5.

[I71 Ruotolo C, Nottage WM. Surgical and nonsurgical management Of rotator Cuff tears. Arthroscopy 2002;18:527-31.

[I81 Silva MJ, Boyer MI, Ditsios K, et al. The insertion site of the canine flexor digitorum profundus tendon heals slowly following injury and suture repair. J Orthop Res 2002;20:447-53.