- Email: [email protected]
Flexor digitorum profundus tendon to bone tunnel repair: A vascularization and histologic study in canines
Flexor Digitorum Profundus Tendon to Bone Tunnel Repair: A Vascularization and Histologic Study in Canines Thomas T. Dovan, MD, Timothy Ritty, PhD, Konstantinos Ditsios, MD, Matthew J. Silva, PhD, Nozomu Kusano, MD, Richard H. Gelberman, MD, St. Louis, MO
Purpose: Recent in vivo canine studies have shown incomplete restoration of the flexor digitorum profundus (FDP) insertion site after transection and repair to the cortical surface of the distal phalanx. Previous biomechanical analyses of tendon to bone surface repair have suggested that repair site gap formation of greater than 3 mm occurs frequently under physiologic loads. A recent ex vivo investigation into a novel repair of the FDP tendon into a bone tunnel in the distal phalanx showed improved tensile properties with a decrease in repair site gap formation. Time-zero data, however, do not always accurately reflect in vivo responses. The repair response of the FDP tendon when placed in an osseous compartment is not known. The purpose of this study was to analyze the histologic and vascular anatomic properties of the FDP insertion site after transection and repair in a bone tunnel within the distal phalanx. Methods: Twenty-six FDP tendon to bone repairs were performed in 13 adult mongrel dogs after insertion site transection. The tendons were repaired in a bone tunnel in the distal phalanx. Vascular analysis of the tendon and repair site was performed by using a modified Spalteholtz technique and routine hematoxylin-eosin staining was used to assess histologic properties of the repair. Results: In normal specimens the vascular analysis showed that there was a distal network of vessels extending 1- to 2-cm proximal to the FDP insertion site. At 10 days after repair the distal tendon segment tendon remained avascular. By 21 days after repair there was proximal migration of an unorganized reticular network of tendon surface vessels with sparse intratendinous communications. At 6 weeks after repair the structure of the distal tendon vascular network resembled that of normals. The vascular response of the tendon within the bone tunnel followed a similar time frame. Histologic analysis showed an inflammatory reaction in the bone tunnel leading to a progressive degradation of that portion of the FDP tendon that resided in the tunnel. Tendon necrosis was not seen. Conclusions: The FDP tendon, after insertion site transection and repair in a bone tunnel, undergoes a process of neovascularization and revascularization over a period of 6 weeks. There is a progressive loss of tendon parenchyma within the bone tunnel and the suture tracks appeared
From the Department of Orthopaedic Surgery, Barnes-Jewish Hospital at Washington University, St. Louis, MO. Received for publication August 8, 2003; accepted in revised form February 8, 2004. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. Supported by the National Institutes of Health (AR33097). Reprint requests: Richard H. Gelberman, MD, One Barnes-Jewish Hospital Plaza, Suite 11300, St. Louis, MO 63110. Copyright © 2005 by the American Society for Surgery of the Hand 0363-5023/05/30A02-0006$30.00/0 doi:10.1016/j.jhsa.2004.02.007
246
The Journal of Hand Surgery
Dovan et al / Canine FDP Tendon to Bone Repair
247
to serve as conduits for the ingrowth of inflammatory tissue. Restoration of the normal 4-zone tendon– bone interface was not seen. Although ex vivo biomechanical assessment of tendon repair in a bone tunnel appears promising, the repair response in vivo may not be favorable for tendon to bone healing. The progressive tendon degeneration that was observed here may have detrimental effects on repair site tensile properties, increasing the potential for early failure. (J Hand Surg 2005;30A:246 –257. Copyright © 2005 by the American Society for Surgery of the Hand.) Key words: Bone tunnel, flexor digitorum profundus, tendon– bone repair, tendon vascularity.
The flexor digitorum profundus (FDP) tendon, when avulsed from bone or transected within zone I, commonly is repaired by direct suture of tendon to bone.1,2 The maintenance of repair is dependent on the ability of the newly constructed tendon– bone interface to withstand the forces generated by postoperative rehabilitation. The available clinical data indicate that patients often have incomplete recovery of motion after injury and repair. Scientific studies have suggested that failure may occur owing to excessive gap formation, suture rupture, or suture pullout from tendon.3–5 In addition a recent in vivo canine study suggested that when the FDP tendon is sutured to the cortical surface of the distal phalanx the bone plays a passive role in the repair process. This observation was supported by a lack of cellular contribution from the distal phalanx to the repair site.4 Observations from studies on tendon to bone repairs in other anatomic regions suggest that the placement of tendon in a bone tunnel may facilitate tissue integration via a broadened surface area of tendon to bone contact, exposure of the tendon to mesenchymal stem cells from the marrow of the distal phalanx, and ease of matrix deposition.6 –11 This study is the second phase in the analysis of a novel method for FDP insertion site repair. The purpose of this study was to evaluate both the vascularization and the histologic responses of the canine FDP tendon repaired in a tunnel in the distal phalanx. Our hypothesis was that the distal FDP tendon would undergo progressive revascularization and that the intramedullary compartment of the distal phalanx would show an active role in tendon to bone healing, evidenced by a contribution of vascular and inflammatory components from the osseous compartment.
Materials and Methods Surgical Technique We performed 26 FDP tendon to bone repairs in 13 adult mongrel dogs (23–27 kg, Covance Research Products, Kalamazoo, MI). The FDP tendons from
the second and fifth digits of each right forelimb were transected from their insertions on the distal phalanges and repaired using a 4-strand technique in a bone tunnel (Fig. 1). The dogs were killed at 10, 21, and 42 days. All procedures were performed in an operating room designated for animals and were approved by our institutional animal studies committee. For surgery each dog was anesthetized using thiopental sodium (0.5 mL/kg intravenously), atropine (0.5 mL), and acepromazine (0.2 mL), and then intubated and maintained on 1% isoflurane. The right forelimb then was shaved, prepared with 10% povidone iodine (Betadine, Purdue Frederick Co., Stamford, CT), and exsanguinated. Surgery was performed under tourniquet control. A midlateral incision was made to expose the insertion of the FDP tendon into the volar base of the proximal aspect of the distal phalanx. Tendons were released directly from their bony insertions by sharp dissection. Through a dorsal longitudinal incision over the distal interphalangeal joint the terminal extensor tendon and dorsal capsule were released, and collateral ligaments were recessed in an effort to decrease the degree of hyperextension of the interphalangeal joint. A 5-mm deep by 4-mm diameter bone tunnel was drilled in the proximal volar aspect of the distal phalanx in the area of the anatomic FDP insertion. We used two 4-0 double-armed sutures (Supramid; S. Jackson Inc, Alexandria, VA) to perform a 4-strand modified Becker technique in the distal FDP tendon stump.3 The first suture was passed initially through the tendon in a plane parallel to the volar–
Figure 1. The distal end of the canine FDP tendon was grasped by using a modified Becker stitch and placed in a tunnel in the distal phalanx. The sutures then were passed through 2 holes in the distal phalanx and tied over the dorsal cortex.
248
The Journal of Hand Surgery / Vol. 30A No. 2 March 2005
Figure 2. Modified Becker technique. (A) Sagittal view of the distal FDP tendon with one arm of the double-armed suture showing multiple passes. During an actual repair the 2 arms of the suture are placed simultaneously. (B) Sagittal view of the distal FDP tendon after placement of one double-armed suture. (C) Coronal view of the distal FDP tendon showing placement of 2 double-armed sutures.
dorsal axis of the tendon midway between the central axis and its radial edge. Thus the suture exited the tendon both at the volar and dorsal aspects of the tendon (in a dorsal–volar plane) 15-mm proximal to the distal end. Subsequently the segment of suture that exited volarly was passed in a dorsal to volar direction entering the tendon 3-mm distal to the previous point of exit (Fig. 2A). Similarly the suture segment that exited from the dorsal aspect of the tendon was passed in a volar to dorsal direction, entering the tendon 3-mm distal to the previous point of exit. The suture strands crossed the tendon on both its external and internal surfaces and after repeated passes grasped the tendon along its radial aspect at a total of 4 points. The suture then was passed through the distal end of the tendon stump (Fig. 2B). The second suture was passed on the ulnar side of the tendon in the same style as the first, completing the 4-strand technique (Fig. 2C). Two needles (Keith needles) then were drilled through the previously created hole in the distal phalanx in an antegrade direction, emerging through the dorsal cortex of the distal phalanx proximal to the nail. The sutures then were passed through the needles and the needles were delivered dorsally. The sutures were tied over the dorsal aspect of the distal phalanx. The sheath was not repaired and the incisions were closed using 4-0 nylon. After surgery the right forelimb was immobilized in a spica cast with the wrist positioned in 70° of flexion. The volar portion of the cast had a removable piece to allow for daily passive motion rehabilitation
of the wrist and digits beginning on the first postoperative day while the dorsal portion of the cast served as an extension block splint. During two 5-minute sessions each day the digits were flexed passively and extended with the wrist flexed at 70°. We have determined previously that this protocol produces approximately 2 mm of FDP excursion and an average peak tendon force of 5 N.12 The dogs were allowed unrestricted cage activities in the cast for the remainder of the time. They were killed at 10, 21, or 42 days after surgery by overdose of pentobarbital (3 dogs at 10 days, 6 dogs at 21 days, and 4 dogs at 42 days). Repaired specimens were taken from the right forelimb and intact normal specimens were taken from the contralateral left forelimb.
Perfusion Studies A total of 20 tendon– bone samples were studied (16 repaired specimens from 12 dogs and 4 normals). The median artery, which is the major artery supplying the dog’s paw, was cannulated in both the experimental (right) and normal (left) limbs. The artery then was injected with 60 cc of full-strength India Ink under firm manual pressure with the tendons in situ before dissection of the forepaw. The adequacy of perfusion was judged by the appearance of India Ink at the freshly cut volar pads of the toes. The entire forelimb then was stored at 4°C for 24 to 48 hours to enhance the fixation of the India Ink in the specimen and to facilitate the final dissection. The distal phalanx, the FDP tendon, and the volar plate were dissected from the forepaw using ⫻2.5 loupe
Dovan et al / Canine FDP Tendon to Bone Repair
magnification and a stereodissection microscope, ⫻6.3 to ⫻32 (Wild Photomakroskop, Heerbrugg, Switzerland). The FDP tendon was transected in the region of the proximal part of the mesotenon, an area that is proximal to the intrasynovial portion of the tendon. The specimens then were pinned at their ends to wooden tongue blades and fixed in 10% formalin for 48 hours followed by a distilled water (dH2O) wash for 2 hours. Decalcification was performed in a 30% formic acid solution changed at 3-day intervals for 4 weeks. After decalcification the specimens were rinsed with dH2O for 12 hours and progressively dehydrated in increasing concentrations of ethanol (30%, 50%, 60%, 75%, 95%) changed at 12-hour intervals. The final dehydration step was 100% ethanol for 24 hours. Tendon and bone clarification was completed by immersion in methyl salicylate. The cleared specimens were stored in methyl salicylate.
General Histology A total of 18 tendon– bone samples were studied (10 repaired specimens from 9 dogs and 8 normals). The tendon– bone repairs were dissected within 1 hour postmortem and fixed in buffered 10% formalin at 4°C for 24 hours. The samples then were decalcified at 4° with 0.5 mol/L ethylenediaminetetraacetic acid for 6 to 8 weeks. During that time the nail surrounding the distal phalanx was removed to facilitate decalcification and subsequent sectioning. After decalcification the samples underwent stepwise dehydration through ethanol and xylene, were embedded in paraffin, and sections were cut at 5-m thickness. Routine hematoxylin-eosin staining was performed. Sections were examined by light microscopy and images were captured with a digital camera.
Results Six of 26 repairs were observed to be ruptured or excessively gapped at necropsy. There were no failures in the 10-day group, 5 failures in the 21-day group, and 1 failure in the 42-day group. Remaining for vascular analysis were 4 tendon– bone repairs at 10, 21, and 42 days, and for histologic analysis were 2, 3, and 2 tendons at 10, 21, and 42 days, respectively.
Normal Specimens Our findings on the normal vascular anatomy of the canine FDP tendon paralleled those of a previous canine study.13 The most proximal region of the FDP tendon was vascularized both dorsally and volarly by a well-developed mesotenon. The transition between
249
the mesotenon and the intrasynovial region of the tendon was defined consistently by the synovial reflection. As the vessels enter the intrasynovial portion of the tendon there was a dramatic decrease in the dorsal vasculature and a volar avascular zone was seen. Both superficial (epitendinous) and intratendinous vessels terminated in microvascular loops approximately 8 to 16 mm distal to the synovial reflection. Distal to this termination there was a 30-mm avascular zone, which led to the terminal 20 mm of vascularized tendon. The distal portion of the FDP tendon, supplied by the vinculum breve, showed fine-caliber dorsal longitudinal vessels that terminated in microvascular loops. In the region of the volar plate there was a well-organized plexus of longitudinal vessels on the volar surface of the tendon that extended proximally to the proximal extent of the volar plate. These epitendinous vessels supplied approximately 50% of the volar tendon via abundant intratendinous penetrations. There were no vessels between the dorsal tendon and the volar plate (Fig. 3A). At the proximal extent of the volar plate the volar epitendinous vessels terminated and the volar aspect of the tendon was devoid of vessels. Dorsally at this level there was a well-organized plexus of longitudinal vessels that extended proximally 10 to 22 mm. In this region the intratendinous vessels were sparse in comparison with the dorsal epitendinous vessels. Hematoxylin-eosin staining of the FDP tendon insertion at the distal phalanx revealed typical distinct transitional zones. Distally the normal tendon was replaced by a short fibrocartilaginous zone, containing mature-appearing chondrocytes. Further distally there was a small zone of calcified fibrocartilage that terminated in the osseous surface of the distal phalanx. Aligned collagen fibers were observed to be present throughout the 2 fibrocartilaginous zones. No vascularity was observed within the transitional regions.
10-Day Specimens The vasculature in the region of the mesotenon was similar to that of the normals. There was a regular branching pattern of longitudinal vessels with a sparse network of intratendinous vessels. Distal to the area of the synovial reflection marked changes were noted. The superficial vessels continued in a longitudinal pattern with regular branching, however, a well-defined epitendinous system of vessels was seen circumferentially. This was in stark contrast to the volar avascular zone identified in the normal
250
The Journal of Hand Surgery / Vol. 30A No. 2 March 2005
Figure 3. Axial sections through the volar plate. (A) Normal specimen showing avascularity of volar plate (VP) and dorsal (straight arrow) region of the tendon (T) adjacent to a well-vascularized volar region (curved arrow). (B) A 42-day specimen showing abundant neovascularization of the volar plate (VP) and the previous dorsal (arrow) avascular zone of the tendon (T).
tendons. These vessels terminated approximately 10 mm distal to the synovial reflection. The distal tendon stump was mostly avascular (Fig. 4A). There were sparse irregularly arranged superficial vessels seen dorsally along the tendon in the area of the volar plate, without intratendi-
nous connections. There was a well-developed array of intratendinous vessels that was arranged around the sutures and in the suture tracts. These vessels appeared to extend from the distal–volar aspect of the volar plate and did not extend beyond the most proximal sutures. There were no epiten-
Dovan et al / Canine FDP Tendon to Bone Repair
251
Figure 4. Anteroposterior views of the distal tendon stump just proximal to the volar plate. (A) A 10-day specimen showing an avascular distal tendon stump (T). (B) A 42-day specimen showing revascularization of the distal extraosseous tendon. The 42-day specimen resembles that of normals with regard to vessel organization and proximal extension (straight arrow).
dinous vessels volarly and the tendon substance was devoid of vessels except in the regions immediately around the suture material. The volar plate remained avascular, similar to that noted in the
normals. There was an increase in surface vessels without intrasubstance penetration into the volar plate. There were no adhesions noted between the volar plate and the tendon. Axial sections through
252
The Journal of Hand Surgery / Vol. 30A No. 2 March 2005
the bone tunnel showed tendon parenchyma inside the bone tunnel. The tendon substance was avascular. The marrow of the distal phalanx had a rich network of vascular channels that were similar in density to the normals. Osseous vessels were seen abutting the tendon surface but there was no penetration of tendon tissue noted. Histologic analysis revealed a robust infiltration of granulation tissue within the bone tunnel without evidence of tendon necrosis (Fig. 5). These samples exhibited a low-grade chronic inflammation with multinucleated cells, macrophages, and mild foreign body response in the region of the suture. Some new bone formation was visible at the edge of the drilled tunnel.
21-Day Specimens In the region of the mesotenon the vessels remained longitudinally arranged, however, the vessels were slightly larger in caliber and more tortuous than that seen at 10 days. The proximal intrasynovial region was similar to that of the 10-day specimens with both dorsal and volar epitendinous vessels with sparse intratendinous penetrations. There was no difference in the length of the proximal intrasynovial vessels compared with the normals. Distally there was an abundant increase in vascularity compared with the 10-day specimens. The majority of this neovascularization was dorsal in the area of the volar plate, extending into the epitenon layer 10-mm proximal to the volar plate. Although intratendinous penetration was noted, these penetrations were sparse. The proximal migration of the intratendinous and superficial vessels was greater at 21 days than it was at 10 days. The proximal extent of the distal vessels was similar to that of the normals but the pattern was one of disorganization. There was an abundant vascular infiltration seen around the suture material to a greater extent than that seen in the 10-day specimens with intratendinous connections to the dorsal superficial vessels. In addition, although the volar surface of the distal tendon had a sparse irregular plexus of superficial vessels, the majority of the epitendinous vessels were seen dorsally between the volar plate and the tendon. The distal aspect of the tendon was adhered to the volar plate firmly. Axial cuts through the bone tunnel showed tendon in the tunnel with sparse intratendinous penetrations from osseous channels. Histologically a vigorous proliferation of granulation tissue within the bone tunnel continued. The rapidly proliferating cells appeared to be infiltrating
the tendon both within the bone tunnel and also external to the tunnel. The infiltration was most apparent surrounding the suture, which appeared to act as a conduit. In comparison with the 10-day samples the inflammatory response was greater at 21 days.
42-Day Specimens The proximal vessels were similar in structure to the 21-day specimens with no significant difference in the extension of vessels distally. Although intrasynovial epitendinous vessels were arranged longitudinally in an organized pattern, the vessels were larger in caliber and more tortuous than that seen in the 21-day specimens. As noted in the 10- and 21-day specimens there was a well-developed plexus of dorsal and volar epitendinous vessels with sparse intratendinous penetrations. At 42 days the distal blood supply was more highly developed and better organized than it had been at 21 days. This was shown by an increase in longitudinal surface vessels with more abundant intratendinous penetrations. The vascular pattern observed was more reflective of that seen in the normal specimens. At the level of the volar plate there was abundant neovascularization dorsally with revascularization seen volarly. In addition the intratendinous penetration of the dorsal superficial vessels was greater and there was significant penetration from the volar epitendinous vessels. The volar avascular zone had marked vascular infiltration from the new volar epitendinous vessels. The volar plate showed significant neovascularization with an abundance of surface vessels dorsally and volarly penetrating into the substance of the volar plate (Fig. 3B). Proximal to the volar plate the superficial and intratendinous vessels began to resemble those of the normals (Fig. 4B). The vessels extended approximately 10-mm proximal to the volar plate, with a well-organized plexus of dorsal longitudinal vessels with multiple transverse interconnections and intratendinous penetration. The volar surface was avascular. Axial sections through the distal phalanx showed that the tunnel was filled with both a combination of tendon and amorphous granulation tissue. The contents of the tunnel were well vascularized. The tendon showed vascular infiltration from osseous channels with multiple intratendinous penetrations. The granulation tissue also showed an abundant vascular supply from adjacent osseous channels (Fig. 6). On histologic examination an acute inflammatory response was noted. The proliferation of inflamma-
Dovan et al / Canine FDP Tendon to Bone Repair
253
Figure 5. Hematoxylin-eosin staining of sagittal paraffin section through the bone tunnel in the distal phalanx at 10 days. (A) Granulation tissue (*) began to penetrate the segment of tendon (T) that was inside the bone tunnel (magnification, ⫻4). The articular surface (A) of the distal phalanx is labeled for orientation. (B) Tendon parenchyma within the bone tunnel showing an infiltration of inflammatory cells (magnification, ⫻10).
tory tissue within the bone tunnel was more extensive and in some instances extended into areas of trabecular bone. Within the bone tunnel the tendon was well infiltrated with granulation tissue and disruption
of the tendon parenchyma was noted (Fig. 7). New bone formation appeared to provide a barrier between the injury zone of the tunnel and the trabecular bone.
254
The Journal of Hand Surgery / Vol. 30A No. 2 March 2005
Figure 6. Axial section through the bone tunnel of a 42-day injection specimen. The tendon has abundant vascular penetrations from adjacent osseous channels (curved arrow). A transitional zone at the tendon bone interface is not seen (straight arrows).
Discussion The purpose of our study was to assess the vascular and histologic response to FDP tendon healing in a bone tunnel. Investigations into the healing properties of FDP tendon insertion site injuries have shown that there is a limited capacity to regenerate the normal structure of the tendon bone interface.4 Although repair site rigidity has been shown to increase over time, with concomitant decreases in strain, the ultimate strength of the insertion site has been shown to remain at baseline levels for the first 6 weeks after injury and repair.5 Furthermore, unpublished data from recent histologic and immunohistochemical studies suggest that FDP tendon to bone surface repairs fail frequently owing to the development of gaps of greater than 3 mm. The defects that develop between the tendon and bone are bridged in a progressive fashion by nondirectional fibrocellular proliferation (Boyer, personal communication). Even in those specimens in which the distal tendon stump was noted to remain opposed to the cortical surface of the distal phalanx, reorganization of the repair site
to a mature transitional zone was not seen. Furthermore there was no evidence that the osseous surface played an active role in the repair process.4 These findings, which may explain the high incidence of tendon to bone repair site failures seen experimentally, suggest that clinical loss of distal interphalangeal joint motion that occurs after insertion site repair may be caused, in part, by repair site elongation. Placement of the FDP tendon into a cancellous bone tunnel in the distal phalanx may facilitate the process of tendon to bone incorporation and obviate problems associated with surface healing. Previous studies have shown that tendon healing in a bone tunnel progresses over time.6,7,9 –11 After an initial inflammatory response at the tendon bone interface there is a deposition of new collagen fibers. Woven bone forms along the tunnel edges and as it matures into lamellar bone it grows into the intervening zone of fibrous tissue and into the tendon itself. The tendon in the tunnel, which remains viable, is populated with normal-appearing fibroblasts.9 Whiston and Walmsey10 noted that although early degeneration
Dovan et al / Canine FDP Tendon to Bone Repair
255
Figure 7. (A) Hematoxylin-eosin staining of a sagittal paraffin section through the bone tunnel in the distal phalanx at 42 days (magnification, ⫻4). The granulation tissue is more prominent and there is a progressive loss of histologically identifiable tendon (T) within the bone tunnel. (B) Marked inflammatory response within the bone tunnel with loss of histologically identifiable tendon (magnification, ⫻10).
occurs, the macrostructure of tendon is maintained as host cells from surrounding bone marrow infiltrate and ultimately replace it with new tendon tissue. Our findings indicate that although the distal
stump of the FDP tendon had sequential neovascularization and revascularization over a 6-week time frame, there were profound differences in the patterns of vascularity seen in the intraosseous and
256
The Journal of Hand Surgery / Vol. 30A No. 2 March 2005
extraosseous tendon segments. Over a period of time there was a replication of the vascular architecture of the extraosseous portion of the tendon, such that by 6 weeks it resembled the vascularization of the normal specimens. In contrast the intraosseous tendon segment was revascularized via a circumferential network of vessels that appeared to accompany the inflammatory response along suture tracks and from adjacent osseous channels. This permeative vascular pattern was similar to the intraosseous inflammatory reaction shown histologically. With regard to the intraosseous response we noted a progressive degeneration of that portion of the tendon that resided within the bone tunnel. Although there was a lack of tendon necrosis there was an infiltration of tendon by granulation tissue and new vessels, primarily along suture tracts. The tendon, not appearing to affect its own repair, appeared to become repopulated with extrinsic cells and vessels. At 6 weeks, as granulation tissue became prominent throughout the tendon, there was a progressive loss of histologically identifiable tendon within the tunnel. This degeneration may affect the mechanical integrity of the repair site because approximately 50% of the distal tendon stump was advanced into the bone tunnel, the region of the profound tendon degradation. A potential cause for the marked tendon degeneration seen in our model is the presence of intratendinous suture within the bone tunnel. Previous models of tendon to bone healing used extraosseous methods of fixation, eliminating the presence of tendon suture in the tunnel itself.6 –11 Our observation that the sutures act as conduits for the ingrowth of granulation tissue is supported by findings from the injected specimens, which showed early vascular proliferation along the tracks of suture material. Taken together these findings indicate that the repair process is a complex one and that the cause of repair site elongation and failure may be multifactorial. The intraosseous inflammatory response, shown to occur progressively over 42 days, may cause a softening of tendon, potentiating early gap formation and suture pullout. Previous studies of tendon to bone surface repairs, showing the detrimental effects of gap formation on insertion site healing, have suggested that the cortical bone surface is involved only passively in the repair process.4 Our tunnel model, which was designed to obviate these issues, created a safety zone whereby 5 mm of tendon tissue was advanced into the medullary space of the distal phalanx. Our findings indicate, however, that the biolog-
ical response within the medullary canal may not be a favorable one for tendon, insofar as the enhancement of tendon to bone response is concerned. Although tendon revascularization was shown, we did not assess the tensile properties of the tendon or the tendon– bone interface and cannot infer that revascularization is associated with an improvement in mechanical properties. Similarly we did not assess whether or not the degradation of the FDP tendon in the bone tunnel had a detrimental effect on the tensile properties of the tendon or tendon to bone interface. In addition our findings are limited to tendon repair in a bone tunnel and do not apply to repair of the FDP tendon to the cortical surface of the distal phalanx. Thus although the bone tunnel repair has certain theoretical benefits over surface repair, we have not proven that the 2 methods lead to different functional results. Future efforts in our laboratories are designed to determine new techniques to improve repair site healing and will investigate not only the histologic and immunohistochemical properties of tendon healing in a bone tunnel, but also the mechanical properties in comparison with tendon to bone surface repairs.
References 1. Bunnell S. Surgery of the Hand. Philadelphia: Lippincott, 1944. 2. Leddy JP, Packer JW. Avulsion of the profundus tendon insertion in athletes. J Hand Surg 1977;2:66 – 69. 3. Boyer MI, Ditsios K, Gelberman RH, Leversedge F, Silva M. Repair of flexor digitorum profundus tendon avulsions from bone: an ex vivo biomechanical analysis. J Hand Surg 2002;27A:594 –598. 4. Boyer MI, Harwood F, Ditsios K, Amiel D, Gelberman RH, Silva MJ. Two-portal repair of canine flexor tendon insertion site injuries: histologic and immunohistochemical characterization of healing during the early postoperative period. J Hand Surg 2003;28A:469 – 474. 5. Silva MJ, Boyer MI, Ditsios K, Burns ME, Harwood FL, Amiel D, 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– 453. 6. Liu SH, Panossian V, al-Shaikh R, Tomin E, Shepherd E, Finerman GA, et al. Morphology and matrix composition during early tendon to bone healing. Clin Orthop 1997;339: 253–260. 7. Rodeo SA, Arnoczky SP, Torzilli PA, Hidaka C, Warren RF. Tendon-healing in a bone tunnel. A biomechanical and histological study in the dog. J Bone Joint Surg 1993;75A: 1795–1803. 8. Rodeo SA, Suzuki K, Deng X-H, Wozney J, Warren RF. Use of recombinant human bone morphogenetic protein-2 to enhance tendon healing in a bone tunnel. Am J Sports Med 1999;27:476 – 488. 9. Weiler A, Hoffmann RFG, Bail HJ, Rehm O, Südkamp NP. Tendon healing in a bone tunnel. Part II: Histologic
Dovan et al / Canine FDP Tendon to Bone Repair analysis after biodegradable interference fit fixation in a model of anterior cruciate ligament reconstruction in sheep. Arthroscopy 2002;18:124 –135. 10. Whiston TB, Walmsley R. Some observations on the reaction of bone and tendon after tunnelling of bone and insertion of tendon. J Bone Joint Surg 1960;42B:377 386. 11. Yamakado K, Kitaoka K, Yamada H, Hashiba K, Nakamura R, Tomita K. The influence of mechanical stress on
257
graft healing in a bone tunnel. Arthroscopy 2002; 18:82–90. 12. Lieber RL, Silva MJ, Amiel D, Gelberman RH. Wrist and digital joint motion produce unique flexor tendon force and excursion in the canine forelimb. J Biomech 1999;32:175– 181. 13. Gelberman RH, Khabie V, Cahill CJ. The revascularization of healing flexor tendons in the digital sheath. A vascular injection study in dogs. J Bone Joint Surg 1991;73A:868 – 881.
Purpose: Recent in vivo canine studies have shown incomplete restoration of the flexor digitorum profundus (FDP) insertion site after transection and repair to the cortical surface of the distal phalanx. Previous biomechanical analyses of tendon to bone surface repair have suggested that repair site gap formation of greater than 3 mm occurs frequently under physiologic loads. A recent ex vivo investigation into a novel repair of the FDP tendon into a bone tunnel in the distal phalanx showed improved tensile properties with a decrease in repair site gap formation. Time-zero data, however, do not always accurately reflect in vivo responses. The repair response of the FDP tendon when placed in an osseous compartment is not known. The purpose of this study was to analyze the histologic and vascular anatomic properties of the FDP insertion site after transection and repair in a bone tunnel within the distal phalanx. Methods: Twenty-six FDP tendon to bone repairs were performed in 13 adult mongrel dogs after insertion site transection. The tendons were repaired in a bone tunnel in the distal phalanx. Vascular analysis of the tendon and repair site was performed by using a modified Spalteholtz technique and routine hematoxylin-eosin staining was used to assess histologic properties of the repair. Results: In normal specimens the vascular analysis showed that there was a distal network of vessels extending 1- to 2-cm proximal to the FDP insertion site. At 10 days after repair the distal tendon segment tendon remained avascular. By 21 days after repair there was proximal migration of an unorganized reticular network of tendon surface vessels with sparse intratendinous communications. At 6 weeks after repair the structure of the distal tendon vascular network resembled that of normals. The vascular response of the tendon within the bone tunnel followed a similar time frame. Histologic analysis showed an inflammatory reaction in the bone tunnel leading to a progressive degradation of that portion of the FDP tendon that resided in the tunnel. Tendon necrosis was not seen. Conclusions: The FDP tendon, after insertion site transection and repair in a bone tunnel, undergoes a process of neovascularization and revascularization over a period of 6 weeks. There is a progressive loss of tendon parenchyma within the bone tunnel and the suture tracks appeared
From the Department of Orthopaedic Surgery, Barnes-Jewish Hospital at Washington University, St. Louis, MO. Received for publication August 8, 2003; accepted in revised form February 8, 2004. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. Supported by the National Institutes of Health (AR33097). Reprint requests: Richard H. Gelberman, MD, One Barnes-Jewish Hospital Plaza, Suite 11300, St. Louis, MO 63110. Copyright © 2005 by the American Society for Surgery of the Hand 0363-5023/05/30A02-0006$30.00/0 doi:10.1016/j.jhsa.2004.02.007
246
The Journal of Hand Surgery
Dovan et al / Canine FDP Tendon to Bone Repair
247
to serve as conduits for the ingrowth of inflammatory tissue. Restoration of the normal 4-zone tendon– bone interface was not seen. Although ex vivo biomechanical assessment of tendon repair in a bone tunnel appears promising, the repair response in vivo may not be favorable for tendon to bone healing. The progressive tendon degeneration that was observed here may have detrimental effects on repair site tensile properties, increasing the potential for early failure. (J Hand Surg 2005;30A:246 –257. Copyright © 2005 by the American Society for Surgery of the Hand.) Key words: Bone tunnel, flexor digitorum profundus, tendon– bone repair, tendon vascularity.
The flexor digitorum profundus (FDP) tendon, when avulsed from bone or transected within zone I, commonly is repaired by direct suture of tendon to bone.1,2 The maintenance of repair is dependent on the ability of the newly constructed tendon– bone interface to withstand the forces generated by postoperative rehabilitation. The available clinical data indicate that patients often have incomplete recovery of motion after injury and repair. Scientific studies have suggested that failure may occur owing to excessive gap formation, suture rupture, or suture pullout from tendon.3–5 In addition a recent in vivo canine study suggested that when the FDP tendon is sutured to the cortical surface of the distal phalanx the bone plays a passive role in the repair process. This observation was supported by a lack of cellular contribution from the distal phalanx to the repair site.4 Observations from studies on tendon to bone repairs in other anatomic regions suggest that the placement of tendon in a bone tunnel may facilitate tissue integration via a broadened surface area of tendon to bone contact, exposure of the tendon to mesenchymal stem cells from the marrow of the distal phalanx, and ease of matrix deposition.6 –11 This study is the second phase in the analysis of a novel method for FDP insertion site repair. The purpose of this study was to evaluate both the vascularization and the histologic responses of the canine FDP tendon repaired in a tunnel in the distal phalanx. Our hypothesis was that the distal FDP tendon would undergo progressive revascularization and that the intramedullary compartment of the distal phalanx would show an active role in tendon to bone healing, evidenced by a contribution of vascular and inflammatory components from the osseous compartment.
Materials and Methods Surgical Technique We performed 26 FDP tendon to bone repairs in 13 adult mongrel dogs (23–27 kg, Covance Research Products, Kalamazoo, MI). The FDP tendons from
the second and fifth digits of each right forelimb were transected from their insertions on the distal phalanges and repaired using a 4-strand technique in a bone tunnel (Fig. 1). The dogs were killed at 10, 21, and 42 days. All procedures were performed in an operating room designated for animals and were approved by our institutional animal studies committee. For surgery each dog was anesthetized using thiopental sodium (0.5 mL/kg intravenously), atropine (0.5 mL), and acepromazine (0.2 mL), and then intubated and maintained on 1% isoflurane. The right forelimb then was shaved, prepared with 10% povidone iodine (Betadine, Purdue Frederick Co., Stamford, CT), and exsanguinated. Surgery was performed under tourniquet control. A midlateral incision was made to expose the insertion of the FDP tendon into the volar base of the proximal aspect of the distal phalanx. Tendons were released directly from their bony insertions by sharp dissection. Through a dorsal longitudinal incision over the distal interphalangeal joint the terminal extensor tendon and dorsal capsule were released, and collateral ligaments were recessed in an effort to decrease the degree of hyperextension of the interphalangeal joint. A 5-mm deep by 4-mm diameter bone tunnel was drilled in the proximal volar aspect of the distal phalanx in the area of the anatomic FDP insertion. We used two 4-0 double-armed sutures (Supramid; S. Jackson Inc, Alexandria, VA) to perform a 4-strand modified Becker technique in the distal FDP tendon stump.3 The first suture was passed initially through the tendon in a plane parallel to the volar–
Figure 1. The distal end of the canine FDP tendon was grasped by using a modified Becker stitch and placed in a tunnel in the distal phalanx. The sutures then were passed through 2 holes in the distal phalanx and tied over the dorsal cortex.
248
The Journal of Hand Surgery / Vol. 30A No. 2 March 2005
Figure 2. Modified Becker technique. (A) Sagittal view of the distal FDP tendon with one arm of the double-armed suture showing multiple passes. During an actual repair the 2 arms of the suture are placed simultaneously. (B) Sagittal view of the distal FDP tendon after placement of one double-armed suture. (C) Coronal view of the distal FDP tendon showing placement of 2 double-armed sutures.
dorsal axis of the tendon midway between the central axis and its radial edge. Thus the suture exited the tendon both at the volar and dorsal aspects of the tendon (in a dorsal–volar plane) 15-mm proximal to the distal end. Subsequently the segment of suture that exited volarly was passed in a dorsal to volar direction entering the tendon 3-mm distal to the previous point of exit (Fig. 2A). Similarly the suture segment that exited from the dorsal aspect of the tendon was passed in a volar to dorsal direction, entering the tendon 3-mm distal to the previous point of exit. The suture strands crossed the tendon on both its external and internal surfaces and after repeated passes grasped the tendon along its radial aspect at a total of 4 points. The suture then was passed through the distal end of the tendon stump (Fig. 2B). The second suture was passed on the ulnar side of the tendon in the same style as the first, completing the 4-strand technique (Fig. 2C). Two needles (Keith needles) then were drilled through the previously created hole in the distal phalanx in an antegrade direction, emerging through the dorsal cortex of the distal phalanx proximal to the nail. The sutures then were passed through the needles and the needles were delivered dorsally. The sutures were tied over the dorsal aspect of the distal phalanx. The sheath was not repaired and the incisions were closed using 4-0 nylon. After surgery the right forelimb was immobilized in a spica cast with the wrist positioned in 70° of flexion. The volar portion of the cast had a removable piece to allow for daily passive motion rehabilitation
of the wrist and digits beginning on the first postoperative day while the dorsal portion of the cast served as an extension block splint. During two 5-minute sessions each day the digits were flexed passively and extended with the wrist flexed at 70°. We have determined previously that this protocol produces approximately 2 mm of FDP excursion and an average peak tendon force of 5 N.12 The dogs were allowed unrestricted cage activities in the cast for the remainder of the time. They were killed at 10, 21, or 42 days after surgery by overdose of pentobarbital (3 dogs at 10 days, 6 dogs at 21 days, and 4 dogs at 42 days). Repaired specimens were taken from the right forelimb and intact normal specimens were taken from the contralateral left forelimb.
Perfusion Studies A total of 20 tendon– bone samples were studied (16 repaired specimens from 12 dogs and 4 normals). The median artery, which is the major artery supplying the dog’s paw, was cannulated in both the experimental (right) and normal (left) limbs. The artery then was injected with 60 cc of full-strength India Ink under firm manual pressure with the tendons in situ before dissection of the forepaw. The adequacy of perfusion was judged by the appearance of India Ink at the freshly cut volar pads of the toes. The entire forelimb then was stored at 4°C for 24 to 48 hours to enhance the fixation of the India Ink in the specimen and to facilitate the final dissection. The distal phalanx, the FDP tendon, and the volar plate were dissected from the forepaw using ⫻2.5 loupe
Dovan et al / Canine FDP Tendon to Bone Repair
magnification and a stereodissection microscope, ⫻6.3 to ⫻32 (Wild Photomakroskop, Heerbrugg, Switzerland). The FDP tendon was transected in the region of the proximal part of the mesotenon, an area that is proximal to the intrasynovial portion of the tendon. The specimens then were pinned at their ends to wooden tongue blades and fixed in 10% formalin for 48 hours followed by a distilled water (dH2O) wash for 2 hours. Decalcification was performed in a 30% formic acid solution changed at 3-day intervals for 4 weeks. After decalcification the specimens were rinsed with dH2O for 12 hours and progressively dehydrated in increasing concentrations of ethanol (30%, 50%, 60%, 75%, 95%) changed at 12-hour intervals. The final dehydration step was 100% ethanol for 24 hours. Tendon and bone clarification was completed by immersion in methyl salicylate. The cleared specimens were stored in methyl salicylate.
General Histology A total of 18 tendon– bone samples were studied (10 repaired specimens from 9 dogs and 8 normals). The tendon– bone repairs were dissected within 1 hour postmortem and fixed in buffered 10% formalin at 4°C for 24 hours. The samples then were decalcified at 4° with 0.5 mol/L ethylenediaminetetraacetic acid for 6 to 8 weeks. During that time the nail surrounding the distal phalanx was removed to facilitate decalcification and subsequent sectioning. After decalcification the samples underwent stepwise dehydration through ethanol and xylene, were embedded in paraffin, and sections were cut at 5-m thickness. Routine hematoxylin-eosin staining was performed. Sections were examined by light microscopy and images were captured with a digital camera.
Results Six of 26 repairs were observed to be ruptured or excessively gapped at necropsy. There were no failures in the 10-day group, 5 failures in the 21-day group, and 1 failure in the 42-day group. Remaining for vascular analysis were 4 tendon– bone repairs at 10, 21, and 42 days, and for histologic analysis were 2, 3, and 2 tendons at 10, 21, and 42 days, respectively.
Normal Specimens Our findings on the normal vascular anatomy of the canine FDP tendon paralleled those of a previous canine study.13 The most proximal region of the FDP tendon was vascularized both dorsally and volarly by a well-developed mesotenon. The transition between
249
the mesotenon and the intrasynovial region of the tendon was defined consistently by the synovial reflection. As the vessels enter the intrasynovial portion of the tendon there was a dramatic decrease in the dorsal vasculature and a volar avascular zone was seen. Both superficial (epitendinous) and intratendinous vessels terminated in microvascular loops approximately 8 to 16 mm distal to the synovial reflection. Distal to this termination there was a 30-mm avascular zone, which led to the terminal 20 mm of vascularized tendon. The distal portion of the FDP tendon, supplied by the vinculum breve, showed fine-caliber dorsal longitudinal vessels that terminated in microvascular loops. In the region of the volar plate there was a well-organized plexus of longitudinal vessels on the volar surface of the tendon that extended proximally to the proximal extent of the volar plate. These epitendinous vessels supplied approximately 50% of the volar tendon via abundant intratendinous penetrations. There were no vessels between the dorsal tendon and the volar plate (Fig. 3A). At the proximal extent of the volar plate the volar epitendinous vessels terminated and the volar aspect of the tendon was devoid of vessels. Dorsally at this level there was a well-organized plexus of longitudinal vessels that extended proximally 10 to 22 mm. In this region the intratendinous vessels were sparse in comparison with the dorsal epitendinous vessels. Hematoxylin-eosin staining of the FDP tendon insertion at the distal phalanx revealed typical distinct transitional zones. Distally the normal tendon was replaced by a short fibrocartilaginous zone, containing mature-appearing chondrocytes. Further distally there was a small zone of calcified fibrocartilage that terminated in the osseous surface of the distal phalanx. Aligned collagen fibers were observed to be present throughout the 2 fibrocartilaginous zones. No vascularity was observed within the transitional regions.
10-Day Specimens The vasculature in the region of the mesotenon was similar to that of the normals. There was a regular branching pattern of longitudinal vessels with a sparse network of intratendinous vessels. Distal to the area of the synovial reflection marked changes were noted. The superficial vessels continued in a longitudinal pattern with regular branching, however, a well-defined epitendinous system of vessels was seen circumferentially. This was in stark contrast to the volar avascular zone identified in the normal
250
The Journal of Hand Surgery / Vol. 30A No. 2 March 2005
Figure 3. Axial sections through the volar plate. (A) Normal specimen showing avascularity of volar plate (VP) and dorsal (straight arrow) region of the tendon (T) adjacent to a well-vascularized volar region (curved arrow). (B) A 42-day specimen showing abundant neovascularization of the volar plate (VP) and the previous dorsal (arrow) avascular zone of the tendon (T).
tendons. These vessels terminated approximately 10 mm distal to the synovial reflection. The distal tendon stump was mostly avascular (Fig. 4A). There were sparse irregularly arranged superficial vessels seen dorsally along the tendon in the area of the volar plate, without intratendi-
nous connections. There was a well-developed array of intratendinous vessels that was arranged around the sutures and in the suture tracts. These vessels appeared to extend from the distal–volar aspect of the volar plate and did not extend beyond the most proximal sutures. There were no epiten-
Dovan et al / Canine FDP Tendon to Bone Repair
251
Figure 4. Anteroposterior views of the distal tendon stump just proximal to the volar plate. (A) A 10-day specimen showing an avascular distal tendon stump (T). (B) A 42-day specimen showing revascularization of the distal extraosseous tendon. The 42-day specimen resembles that of normals with regard to vessel organization and proximal extension (straight arrow).
dinous vessels volarly and the tendon substance was devoid of vessels except in the regions immediately around the suture material. The volar plate remained avascular, similar to that noted in the
normals. There was an increase in surface vessels without intrasubstance penetration into the volar plate. There were no adhesions noted between the volar plate and the tendon. Axial sections through
252
The Journal of Hand Surgery / Vol. 30A No. 2 March 2005
the bone tunnel showed tendon parenchyma inside the bone tunnel. The tendon substance was avascular. The marrow of the distal phalanx had a rich network of vascular channels that were similar in density to the normals. Osseous vessels were seen abutting the tendon surface but there was no penetration of tendon tissue noted. Histologic analysis revealed a robust infiltration of granulation tissue within the bone tunnel without evidence of tendon necrosis (Fig. 5). These samples exhibited a low-grade chronic inflammation with multinucleated cells, macrophages, and mild foreign body response in the region of the suture. Some new bone formation was visible at the edge of the drilled tunnel.
21-Day Specimens In the region of the mesotenon the vessels remained longitudinally arranged, however, the vessels were slightly larger in caliber and more tortuous than that seen at 10 days. The proximal intrasynovial region was similar to that of the 10-day specimens with both dorsal and volar epitendinous vessels with sparse intratendinous penetrations. There was no difference in the length of the proximal intrasynovial vessels compared with the normals. Distally there was an abundant increase in vascularity compared with the 10-day specimens. The majority of this neovascularization was dorsal in the area of the volar plate, extending into the epitenon layer 10-mm proximal to the volar plate. Although intratendinous penetration was noted, these penetrations were sparse. The proximal migration of the intratendinous and superficial vessels was greater at 21 days than it was at 10 days. The proximal extent of the distal vessels was similar to that of the normals but the pattern was one of disorganization. There was an abundant vascular infiltration seen around the suture material to a greater extent than that seen in the 10-day specimens with intratendinous connections to the dorsal superficial vessels. In addition, although the volar surface of the distal tendon had a sparse irregular plexus of superficial vessels, the majority of the epitendinous vessels were seen dorsally between the volar plate and the tendon. The distal aspect of the tendon was adhered to the volar plate firmly. Axial cuts through the bone tunnel showed tendon in the tunnel with sparse intratendinous penetrations from osseous channels. Histologically a vigorous proliferation of granulation tissue within the bone tunnel continued. The rapidly proliferating cells appeared to be infiltrating
the tendon both within the bone tunnel and also external to the tunnel. The infiltration was most apparent surrounding the suture, which appeared to act as a conduit. In comparison with the 10-day samples the inflammatory response was greater at 21 days.
42-Day Specimens The proximal vessels were similar in structure to the 21-day specimens with no significant difference in the extension of vessels distally. Although intrasynovial epitendinous vessels were arranged longitudinally in an organized pattern, the vessels were larger in caliber and more tortuous than that seen in the 21-day specimens. As noted in the 10- and 21-day specimens there was a well-developed plexus of dorsal and volar epitendinous vessels with sparse intratendinous penetrations. At 42 days the distal blood supply was more highly developed and better organized than it had been at 21 days. This was shown by an increase in longitudinal surface vessels with more abundant intratendinous penetrations. The vascular pattern observed was more reflective of that seen in the normal specimens. At the level of the volar plate there was abundant neovascularization dorsally with revascularization seen volarly. In addition the intratendinous penetration of the dorsal superficial vessels was greater and there was significant penetration from the volar epitendinous vessels. The volar avascular zone had marked vascular infiltration from the new volar epitendinous vessels. The volar plate showed significant neovascularization with an abundance of surface vessels dorsally and volarly penetrating into the substance of the volar plate (Fig. 3B). Proximal to the volar plate the superficial and intratendinous vessels began to resemble those of the normals (Fig. 4B). The vessels extended approximately 10-mm proximal to the volar plate, with a well-organized plexus of dorsal longitudinal vessels with multiple transverse interconnections and intratendinous penetration. The volar surface was avascular. Axial sections through the distal phalanx showed that the tunnel was filled with both a combination of tendon and amorphous granulation tissue. The contents of the tunnel were well vascularized. The tendon showed vascular infiltration from osseous channels with multiple intratendinous penetrations. The granulation tissue also showed an abundant vascular supply from adjacent osseous channels (Fig. 6). On histologic examination an acute inflammatory response was noted. The proliferation of inflamma-
Dovan et al / Canine FDP Tendon to Bone Repair
253
Figure 5. Hematoxylin-eosin staining of sagittal paraffin section through the bone tunnel in the distal phalanx at 10 days. (A) Granulation tissue (*) began to penetrate the segment of tendon (T) that was inside the bone tunnel (magnification, ⫻4). The articular surface (A) of the distal phalanx is labeled for orientation. (B) Tendon parenchyma within the bone tunnel showing an infiltration of inflammatory cells (magnification, ⫻10).
tory tissue within the bone tunnel was more extensive and in some instances extended into areas of trabecular bone. Within the bone tunnel the tendon was well infiltrated with granulation tissue and disruption
of the tendon parenchyma was noted (Fig. 7). New bone formation appeared to provide a barrier between the injury zone of the tunnel and the trabecular bone.
254
The Journal of Hand Surgery / Vol. 30A No. 2 March 2005
Figure 6. Axial section through the bone tunnel of a 42-day injection specimen. The tendon has abundant vascular penetrations from adjacent osseous channels (curved arrow). A transitional zone at the tendon bone interface is not seen (straight arrows).
Discussion The purpose of our study was to assess the vascular and histologic response to FDP tendon healing in a bone tunnel. Investigations into the healing properties of FDP tendon insertion site injuries have shown that there is a limited capacity to regenerate the normal structure of the tendon bone interface.4 Although repair site rigidity has been shown to increase over time, with concomitant decreases in strain, the ultimate strength of the insertion site has been shown to remain at baseline levels for the first 6 weeks after injury and repair.5 Furthermore, unpublished data from recent histologic and immunohistochemical studies suggest that FDP tendon to bone surface repairs fail frequently owing to the development of gaps of greater than 3 mm. The defects that develop between the tendon and bone are bridged in a progressive fashion by nondirectional fibrocellular proliferation (Boyer, personal communication). Even in those specimens in which the distal tendon stump was noted to remain opposed to the cortical surface of the distal phalanx, reorganization of the repair site
to a mature transitional zone was not seen. Furthermore there was no evidence that the osseous surface played an active role in the repair process.4 These findings, which may explain the high incidence of tendon to bone repair site failures seen experimentally, suggest that clinical loss of distal interphalangeal joint motion that occurs after insertion site repair may be caused, in part, by repair site elongation. Placement of the FDP tendon into a cancellous bone tunnel in the distal phalanx may facilitate the process of tendon to bone incorporation and obviate problems associated with surface healing. Previous studies have shown that tendon healing in a bone tunnel progresses over time.6,7,9 –11 After an initial inflammatory response at the tendon bone interface there is a deposition of new collagen fibers. Woven bone forms along the tunnel edges and as it matures into lamellar bone it grows into the intervening zone of fibrous tissue and into the tendon itself. The tendon in the tunnel, which remains viable, is populated with normal-appearing fibroblasts.9 Whiston and Walmsey10 noted that although early degeneration
Dovan et al / Canine FDP Tendon to Bone Repair
255
Figure 7. (A) Hematoxylin-eosin staining of a sagittal paraffin section through the bone tunnel in the distal phalanx at 42 days (magnification, ⫻4). The granulation tissue is more prominent and there is a progressive loss of histologically identifiable tendon (T) within the bone tunnel. (B) Marked inflammatory response within the bone tunnel with loss of histologically identifiable tendon (magnification, ⫻10).
occurs, the macrostructure of tendon is maintained as host cells from surrounding bone marrow infiltrate and ultimately replace it with new tendon tissue. Our findings indicate that although the distal
stump of the FDP tendon had sequential neovascularization and revascularization over a 6-week time frame, there were profound differences in the patterns of vascularity seen in the intraosseous and
256
The Journal of Hand Surgery / Vol. 30A No. 2 March 2005
extraosseous tendon segments. Over a period of time there was a replication of the vascular architecture of the extraosseous portion of the tendon, such that by 6 weeks it resembled the vascularization of the normal specimens. In contrast the intraosseous tendon segment was revascularized via a circumferential network of vessels that appeared to accompany the inflammatory response along suture tracks and from adjacent osseous channels. This permeative vascular pattern was similar to the intraosseous inflammatory reaction shown histologically. With regard to the intraosseous response we noted a progressive degeneration of that portion of the tendon that resided within the bone tunnel. Although there was a lack of tendon necrosis there was an infiltration of tendon by granulation tissue and new vessels, primarily along suture tracts. The tendon, not appearing to affect its own repair, appeared to become repopulated with extrinsic cells and vessels. At 6 weeks, as granulation tissue became prominent throughout the tendon, there was a progressive loss of histologically identifiable tendon within the tunnel. This degeneration may affect the mechanical integrity of the repair site because approximately 50% of the distal tendon stump was advanced into the bone tunnel, the region of the profound tendon degradation. A potential cause for the marked tendon degeneration seen in our model is the presence of intratendinous suture within the bone tunnel. Previous models of tendon to bone healing used extraosseous methods of fixation, eliminating the presence of tendon suture in the tunnel itself.6 –11 Our observation that the sutures act as conduits for the ingrowth of granulation tissue is supported by findings from the injected specimens, which showed early vascular proliferation along the tracks of suture material. Taken together these findings indicate that the repair process is a complex one and that the cause of repair site elongation and failure may be multifactorial. The intraosseous inflammatory response, shown to occur progressively over 42 days, may cause a softening of tendon, potentiating early gap formation and suture pullout. Previous studies of tendon to bone surface repairs, showing the detrimental effects of gap formation on insertion site healing, have suggested that the cortical bone surface is involved only passively in the repair process.4 Our tunnel model, which was designed to obviate these issues, created a safety zone whereby 5 mm of tendon tissue was advanced into the medullary space of the distal phalanx. Our findings indicate, however, that the biolog-
ical response within the medullary canal may not be a favorable one for tendon, insofar as the enhancement of tendon to bone response is concerned. Although tendon revascularization was shown, we did not assess the tensile properties of the tendon or the tendon– bone interface and cannot infer that revascularization is associated with an improvement in mechanical properties. Similarly we did not assess whether or not the degradation of the FDP tendon in the bone tunnel had a detrimental effect on the tensile properties of the tendon or tendon to bone interface. In addition our findings are limited to tendon repair in a bone tunnel and do not apply to repair of the FDP tendon to the cortical surface of the distal phalanx. Thus although the bone tunnel repair has certain theoretical benefits over surface repair, we have not proven that the 2 methods lead to different functional results. Future efforts in our laboratories are designed to determine new techniques to improve repair site healing and will investigate not only the histologic and immunohistochemical properties of tendon healing in a bone tunnel, but also the mechanical properties in comparison with tendon to bone surface repairs.
References 1. Bunnell S. Surgery of the Hand. Philadelphia: Lippincott, 1944. 2. Leddy JP, Packer JW. Avulsion of the profundus tendon insertion in athletes. J Hand Surg 1977;2:66 – 69. 3. Boyer MI, Ditsios K, Gelberman RH, Leversedge F, Silva M. Repair of flexor digitorum profundus tendon avulsions from bone: an ex vivo biomechanical analysis. J Hand Surg 2002;27A:594 –598. 4. Boyer MI, Harwood F, Ditsios K, Amiel D, Gelberman RH, Silva MJ. Two-portal repair of canine flexor tendon insertion site injuries: histologic and immunohistochemical characterization of healing during the early postoperative period. J Hand Surg 2003;28A:469 – 474. 5. Silva MJ, Boyer MI, Ditsios K, Burns ME, Harwood FL, Amiel D, 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– 453. 6. Liu SH, Panossian V, al-Shaikh R, Tomin E, Shepherd E, Finerman GA, et al. Morphology and matrix composition during early tendon to bone healing. Clin Orthop 1997;339: 253–260. 7. Rodeo SA, Arnoczky SP, Torzilli PA, Hidaka C, Warren RF. Tendon-healing in a bone tunnel. A biomechanical and histological study in the dog. J Bone Joint Surg 1993;75A: 1795–1803. 8. Rodeo SA, Suzuki K, Deng X-H, Wozney J, Warren RF. Use of recombinant human bone morphogenetic protein-2 to enhance tendon healing in a bone tunnel. Am J Sports Med 1999;27:476 – 488. 9. Weiler A, Hoffmann RFG, Bail HJ, Rehm O, Südkamp NP. Tendon healing in a bone tunnel. Part II: Histologic
Dovan et al / Canine FDP Tendon to Bone Repair analysis after biodegradable interference fit fixation in a model of anterior cruciate ligament reconstruction in sheep. Arthroscopy 2002;18:124 –135. 10. Whiston TB, Walmsley R. Some observations on the reaction of bone and tendon after tunnelling of bone and insertion of tendon. J Bone Joint Surg 1960;42B:377 386. 11. Yamakado K, Kitaoka K, Yamada H, Hashiba K, Nakamura R, Tomita K. The influence of mechanical stress on
257
graft healing in a bone tunnel. Arthroscopy 2002; 18:82–90. 12. Lieber RL, Silva MJ, Amiel D, Gelberman RH. Wrist and digital joint motion produce unique flexor tendon force and excursion in the canine forelimb. J Biomech 1999;32:175– 181. 13. Gelberman RH, Khabie V, Cahill CJ. The revascularization of healing flexor tendons in the digital sheath. A vascular injection study in dogs. J Bone Joint Surg 1991;73A:868 – 881.