Limb Lengthening for Stature: Another View

Abstract

Erratum Noninvasive three-dimensional imaging techniques such as computed tomography (CT) and magnetic resonance have broadened the understanding of the craniofacial skeleton macroarchitecture. They are important tools to analyze gross structural bone changes for pathologic diagnosis and after skeletal surgery. These techniques have been applied to analyze the surgical site after distraction osteogenesis of the mandible, as well as to analyze distraction effects at distant sites such as the temporomandibular joints (TMJs). Currently there is a significant interest in imaging bone at a microscopic resolution to delineate the microarchitecture of cortical and trabecular bone. In the articular regions the bone mainly withstands compressive forces, being entirely trabecular, with a thin subchondral shell of compact bone. The trabeculae of cancellous bone, though individually small, collectively provide powerful support to the surrounding compact bone. Trabecular architecture seems to be optimally structured for its load-bearing function, suggesting that mechanical forces govern its formation. Therefore, bone strength can be greatly influenced not only by bone mass but also by morphologic features such as trabecular connectivity and the architecture of trabecular elements. Recent advances in the application of high-resolution micro-computed tomography (μCT) imaging have been instrumental in providing true quantitative and qualitative three-dimensional data on baseline bone morphology, as well as structural changes that result from load stimulation or lack thereof. It has been hypothesized that the magnitude of the loads imposed on bone dictates its mineralization and structural design. However, it is not known what the long-term effect of bone lengthening is on the articular cartilage of limbs undergoing lengthening for short stature, and whether these joints will last a lifetime. Because distraction osteogenesis has been applied to the mandible and this bone has two joints, it is important to elucidate the possible effects of mandibular lengthening on the TMJs. The biomechanical environment of both TMJs is altered during mandibular distraction osteogenesis. Previous studies involving unilateral linear distraction with external devices have shown temporary changes with no long-term detrimental effects on the temporomandibular joints. Osteoclastic activity in the TMJ has been reported, but changes seem to be partly reversible. There is evidence that bilateral mandibular ramus distraction with an added transverse component may induce condylar erosions similar to those observed in osteoarthritis. Most studies have been conducted at the histologic and macroarchitectural level,and changes in the three-dimensional trabecular bone microstructure of the TMJ following distraction osteogenesis have not been investigated. The aim of this pilot study was to quantify the three-dimensional trabecular bone architectural changes within the glenoid fossa and mandibular condyle in response to bilateral mandibular ramus distraction using μCT in a canine model. Our objective was to better understand the effects of distraction osteogenesis on the glenoid fossa and condylar bone structure after 1- and 2-month consolidation periods. The Institutional Animal Care and Use Committee approved the use and care of animals in this study. Nine 6- to 8-month-old male beagle dogs, weighing between 10 and 15.4 kg, were housed for a 5-day period to become acclimated to the research facility. Preoperative and postoperative care of these animals was overseen by veterinarians and husbandry staff to ensure proper humane treatment. The animals were weighed preoperatively, the day following surgery, and just before sacrifice. Eight animals were operated on under intravenous pentobarbital (25 mg/kg) and general, halothane-induced, endotracheal anesthesia. Intraoral preparation was performed using half-strength hydrogen peroxide diluted with normal saline solution. The surgical area was shaved and draped in a standard surgical fashion and cleaned with 70% isopropyl alcohol followed by 10% povidone-iodine solution and quartered with sterile towels. The mandible was exposed on each respective side through 6- to 8-cm incisions made parallel and 0.5-cm incisions above the right and left inferior border of the mandible. The masseter muscle was reflected anteriorly, and the mandible was exposed in the supraperiosteal plane. The osteotomy was made with a reciprocating saw, copious saline irrigation, and manual tapping with an osteotome along a line below the mandibular angular process posteriorly and a point posterior to the molars on the anterior aspect of the vertical ramus (). Special attention was paid to avoid injury of the inferior alveolar neurovascular bundle. Internal distraction devices (KLS L.P. Zurich Distractor, Cloverleaf Plate, 15 mm, Martin; Jacksonville, FL) were fixed to the mandible following a plane parallel to the canine mandibular ramus using self-tapping 1.5-mm diameter and 5- to 7-mm long screws to secure the footplates of the device above and below the line of osteotomy. The wound was closed in layers with absorbable sutures in the deep layers and nylon skin sutures in an interrupted fashion. The nylon sutures were removed 7 days postoperatively. All animals were placed on a soft mechanical diet and water ad libitum beginning on the 2nd postoperative day. The dogs received intramuscular ampicillin, 250 mg/mL, twice a day for five days and antibiotic coverage and buprenorphine, 0.3 mg/mL, twice a day for 5 days postoperatively to minimize discomfort or pain. After a latency period of 7 days, distraction was initiated at a rate of 1 mm/d (0.5 mm every 12 hours) for 12 consecutive days. The animals were divided into two experimental groups (N = 4) and killed under anesthesia with intravenous KCL (5 mL; 0.3 mg/mL) after a 30- or 60-day period of consolidation. One unoperated animal served as control and was killed in a similar fashion. The length of the achieved distraction was measured postmortem with an occlusal radiograph taken of the resected hemimandible. An average of 10 mm elongation of the ramus bilaterally was recorded. The TMJs (glenoid fossa and condyle) were carefully harvested en bloc. The specimens were stored in 10% buffered formalin solution. 5+ images The TMJs were imaged at 15-μm isotropic voxels, using a Scanco Model 40 μCT scanner (Wayne, PA). In order to place the specimens consistently inside the 30-mm Perspex holder, a special jig was customized with acrylic and 0.32-inch stainless steel orthodontic wire. The device had four specially designed clasps partially embedded in an acrylic circle base that fit tightly inside the holder. The clasps were adjusted with an orthodontic plier to hold the TMJ specimens steady by the glenoid fossa and its portion of the zygomatic arch (). The acrylic base had red ink marks that divided it equally into four segments. These marks served as reference points to position the TMJ specimens with the same orientation. Vertical lines on the lateral side of the holder also assisted in vertically positioning the specimens. In each TMJ, two segments of orthodontic ligature wire (0.012 inch) were placed parallel to the base of the tube and around the most superior and inferior aspect of the area to be scanned. Therefore, from the μCT scout view, the volume of interest of each TMJ could be identified and standardized for all specimens. The anatomic region comprising from the superior to the medial aspect of the TMJ was chosen to be scanned as this was considered the region submitted to more pressure due to the vector of the forces applied by the distractor. Approximately 500 contiguous slices per specimen were generated. From this volume, comparable regions of interest in the condyles and the glenoid fossa were selected and approximately 200 slices were used for trabecular three-dimensional reconstruction. None of these slices included the orthodontic ligature wire. The gray-scale images were processed using a low-pass filter to remove noise, and a fixed threshold of 205 was used to separate the bone from the marrow. From the resulting binarized three-dimensional reconstruction, Scanco software was used to calculate the bone volume per tissue volume (BV/TV) ratio of cancellous bone volume in the total tissue volume (volume of interest); mean trabecular thickness (Tb.Th.); mean trabecular separation (Tb.Sp.); mean trabecular number (Tb.N); structural model index (SMI) - varying from zero to 3 for ideal plate and rod structures, respectively; connectivity density (Conn.D), defined as a measure of the degree to which the structure of the trabecular bone is multiply connected; and anisotropy (DA), the degree of asymmetry in trabecular bone orientation (highly oriented > 1 to less oriented = 1). Cortical bone thickness was visually inspected to compare differences among 1-month, 2-month, and control specimens using three μCT slices selected from each glenoid fossa and condyle. 5+ images Trabecular bone microarchitecture measurements included areas of interest from 36 bones (16 glenoid fossas and 16 condyles divided equally in the two experimental groups and two glenoid fossas and two condyles from an unoperated control dog). The data were analyzed using a commercially available software package (SPSS, Chicago, IL). A paired sample t-test was run to analyze differences in the trabecular bone microarchitecture between the left and right side of each dog. There was no difference between sides so the average of the two sides was calculated and is reported here. Differences in trabecular bone microarchitecture between 1 and 2 months were analyzed by using an independent sample t-test. A one-sample t-test was used to compare each experimental group with the control. P values less than 0.05 were considered significant, but exact values are given. Body weight had changed slightly throughout the experiment in both groups (). Symmetric protrusion, extrusion of the molars, and open bite developed bilaterally after distraction (). Examination of μCT slices indicated that the control group had thicker cortical bone and less prominent trabecular spaces when compared with corresponding sites of the distracted animals (). The three-dimensional reconstructions of the trabecular bone of the glenoid fossa and the condyle indicated that the control specimen had denser bone that was more platelike than the two experimental groups (). 5+ images Seven morphologic parameters were analyzed to assess bone quantity (BV/TV) and quality (Tb.Th, Tb.Sp, Tb.N., Conn.D., SMI, DA) with the use of μCT. These parameters are described first for the glenoid fossa () and then for the condyle (). 5+ images Control Versus 1 Month Significantly lower values were observed for BV/TV (P = 0.028), Tb.N. (P = 0.009), and Conn.D. (P = 0.002), whereas Tb.Sp. (P = 0.032) was significantly higher in the 1-month group compared with the control. DA (P = 0.05) tended to be less and SMI (P = 0.071) tended to be greater than in the control, but the differences were not statistically significant. Control Versus 2 months Tb.Th. (P = 0.008) was smaller and SMI (P = 0.033) was greater in the 2-month group than the control. BV/TV (P = 0.079) tended to be below control levels but the difference was not statistically significant. 1 Versus 2 Months No significant differences were observed in the glenoid fossa trabecular parameters between the 1- and 2-month specimens. Values for the mandibular condyles are shown in (). Control Versus 1 Month At 1 month, there were significantly lower values for BV/TV (P = 0.04), Tb.N (P = 0.030), and Conn.D (P = 0.017) than the control, with a trend in the same direction for Tb.Th (P = 0.053). Tb.Sp. (P = 0.079) tended to be greater, but the difference was not statistically significant. Control Versus 2 Months BV/TV (P = 0.001), Tb.N (P = 0.030), Tb.Th (P = 0.00), and Conn.D (P = 0.011) had significantly lower values and SMI (P = 0.002) was significantly higher at 2 months than the control. 1 Versus 2 Months No significant differences were observed in the condylar trabecular parameters between the 1- and 2-month specimens. In this study the trabecular architecture of the canine glenoid fossa and mandibular condyle was quantitatively evaluated with μCT after 1 and 2 months of consolidation following vertical ramus mandibular distraction osteogenesis. Following distraction osteogenesis, bone volume was decreased compared with the control. This decrease in bone density was due to decreased trabecular thickness and number, with an increase in trabecular spacing. Consequently the trabecular architecture became more rodlike. Thus, our observations indicate that distraction osteogenesis of the mandible induces changes in bone architecture at the TMJ. It is not known whether these changes are transient or permanent. It has been shown that remodeling of the TMJ, in the form of bone contour changes on both the glenoid fossa and condyle, occurs as a response to bilateral mandibular sagittal osteotomies in primates. Bell and Kennedy indicated that the vessels that perfuse the condyle emerge from the lateral pterygoid muscle and the joint capsule. These structures were not disturbed by the surgical procedure in this study. However, the masseter muscle was detached from the mandible to allow fixation of the semiburied distractor device. It has been demonstrated that detachment of the rat masseter muscle alone can lead to significant changes in the morphology of the condyle. In addition, Liu et al have demonstrated that relative loss of masseter function could happen as a result of overstretching after excessive distraction, possibly leading to a decrease in density and size of the condyle in rats. Although the difference between trabecular architecture at 1 and 2 months was not statistically significant, the values after 2 months of consolidation tended to be closer to control values than after 1 month of consolidation. This was especially true for the glenoid fossa. This trend was much less evident in the mandibular condyle. This apparent difference in dynamic changes on the two sides of the temporal mandible joint may be explained by the fact that the temporal bone was not directly affected by surgery (e.g., osteotomy, muscle detachment, alteration in blood supply) and the distraction process, whereas the mandible was directly affected. Conn.D and SMI indexes are known to be helpful in the assessment of the three-dimensional nature of bone structures and architectural features indicative of bone quality as they influence the strength of bone similar to what the strategic placement of beams and rafters does for building a structure. With the exception of the glenoid fossa at 2 months, Conn. D. was below the control value. Kinney and Ladd have suggested that recovery of mechanical function might depend on the preservation or restoration of trabecular connectivity. However, it has also been suggested that loss of structural connectivity appears to be an irreversible process. Longer-term studies are needed to determine whether this aspect of trabecular architecture returns to normal. SMI in both glenoid fossa and condyle specimens demonstrated trends of change from platelike toward a rodlike structure. Trabecular structural changes from platelike to rodlike gradually occur with aging. Previous studies have demonstrated a positive correlation between bone strength and parameters such as bone volume, trabecular number, trabecular thickness, and connectivity. Thus, although the mechanical properties of the specimens were not measured in our study, the observed changes in cancellous bone structure imply a decrease in bone strength. In this study, the cancellous bone in the control mandibular condyle and glenoid fossa specimens was anisotropic. Giesen et al observed that the trabecular bone in human mandibular condyles is also somewhat anisotropic where trabecular orientation consists of parallel plates perpendicular to the mediolateral aspect of the condylar axis. Such a trabecular structure is optimally adapted to sustain loads in directions that coincide with the majority of joint forces applied to the condyle during mastication and also to supply nutrition to the avascular cartilage. Despite limited masticatory function after surgery and distraction, the degree of anisotropy did not change in the current study. Thus, despite changes in bone volume and other parameters, anisotropy was not affected by distraction osteogenesis. Similarly, Giesen et al have shown that the degree of trabecular anisotropy did not differ among loaded and less loaded human mandibular condyles, although decreased mechanical load was associated with reduced density, stiffness, and strength. In the current study, most of the indices analyzed suggested trends in trabecular architecture and volume similar to what has been observed in experiments in which bone was aging or normal function impaired. Due to the short time frame of the study and the use of skeletally immature animals, aging probably had no influence on the bone changes. Therefore, it is likely that distraction osteogenesis-induced changes explain the observations. Without baseline controls, it is not possible to know whether altered growth or remodeling kinetics, or both, account for the structural differences found between the experimental animals and the intact control. Significant variation in animal weight before and after surgery was not detected, suggesting that growth of the animals was not impaired, and all animals, including the control, had a similar soft diet. Thus, it is likely that either the surgery, its consequences on the blood supply, or distraction osteogenesis-induced alterations in the biomechanical environment of the TMJ led to the observed changes in bone architecture. A limitation of an animal model is that the surgical procedure is tested in healthy individuals, as opposed to the clinical situation of treating a disease. Biomechanically, mandibular distraction osteogenesis in a previously normally functioning animal probably generates abnormally high forces and pressures at the TMJ. In addition, experimental mandibular distraction osteogenesis induces changes in normal masticatory function, as opposed to the clinical situation, in which distraction osteogenesis leads to more normal mastication. Despite these limitations, the present data indicate that normal loading of the TMJ and normal physical function are critical to maintaining trabecular microarchitecture and bone volume. A second limitation of our study is that we did not include analysis of the soft tissues, including TMJ ligaments, articular capsule, and disk. Changes in these structures following distraction osteogenesis need to be explored as it is possible that patients may be more likely to suffer clinical symptoms if changes occur in the soft tissue components of the joint than in the bone itself. As a consequence of the surgery itself or because of alterations in the biomechanics of the TMJ following distraction osteogenesis, there was decreased bone quantity and quality of the skeletal TMJ structures in the current study. Thus, deterioration of trabecular bone microarchitecture could occur clinically as distraction osteogenesis would induce change in function, even if it is a change toward more normal function. Whether these changes in bone microarchitecture are reversible, have influence on long-term stability of distracted cases, or cause TMJ disorders was not addressed in this study but merit attention clinically. Healthy subchondral trabecular bone is important not only because it deforms during loading, absorbing forces that could destroy the cartilage layer, but also because the vascularization of this area serves, along with synovial fluid, as a source of nutrition for the avascular cartilage. It has been shown that mandibular condylar resorption and condylar atrophy can be a potential undesirable outcome of acute mandibular advancement, leading to relapse after conventional orthognatic surgery. Although gradual increase of mandibular length with distraction could, theoretically, minimize these deleterious effects on the TMJ, the current study demonstrates that distraction osteogenesis at a clinically used rate did affect bone microarchitecture in the TMJ. Therefore, long-term experiments and clinical follow-up are needed for patients who have had distraction osteogenesis to determine whether change in bone microarchitecture is an adaptive response without significant structural and functional complications. Dr. Sant'Anna is a recipient of a scholarship from Capes Agency, Ministry of Education, Brazil. KLS-Martin, Jacksonville, FL. Research grant to Drs. Polley and Figueroa. Grainger Foundation and NIH RR16631 support for Dr. Sumner. Journal of Craniofacial Surgery. 17(3):611-619, May 2006.