| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, UCSD Medical Center, 200 W. Arbor Dr., San Diego, CA
92103.
2
National Orthopedic Imaging Associates, San Mateo Imaging, 715 N. San Mateo
Dr., San Mateo, CA 94401.
3
Klinik Fur Radiologische Diagnostik, Christians-Albrects Universitat, Arnold
Heller Str. 9, 24105 Kiel, Germany.
4
Department of Orthopedics, UCSD Medical Center, San Diego, CA 92103.
5
Department of Radiology, Mail Code 9114, San Diego VA Medical Center, 3500 La
Jolla Village Dr., San Diego, CA 92161.
Received March 8, 2000;
accepted after revision June 2, 2000.
Supported in part by Johnson & Johnson Professional, Raynham, MA.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. One hundred seventy-five 1.3-mm biodegradable pins made of polydioxanone were used to secure a total of 59 osteochondral allografts of the knee. Patients with the pins underwent scanning on a 1.5-T unit with 3.3- to 4-mm contiguous T1-weighted spin-echo (TR/TE, 600/15), fat-saturated proton density-weighted (3000/40), T2-weighted fast spin-echo (3000/63), and three-dimensional spoiled gradient-recalled (47/7; flip angle, 60°) sequences at 3, 6, 12, 24, or 36 months after surgery. Eighty-nine pins were imaged on multiple occasions. Two osteoradiologists interpreted the MR examinations.
RESULTS. More than 80% of the pin channels were visible at 3 and at 6 months after surgery. By 24 months, only 20% of the pin channels were visible, with the remainder having been replaced by bone. At 3 months, nearly 40% of the pins were associated with adjacent marrow edema. Edema generally diminished, involving less than 20% of pins at later time points. Focal cartilage defects were evident at 32% of the pin insertion sites during the first 6 months, but these defects were present in only 4% of the insertion sites thereafter.
CONCLUSION. Biodegradable polydioxanone pins usually resorb completely by 24 months. Marrow edema, presumably representing inflammation related to pin resorption, is infrequent and tends to resolve. Cartilage defects related to pin placement heal spontaneously.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Over the last 15 years, solid implants made from absorbable polymers have been developed as a potential solution. Reportedly, these implants are resorbed within bone tissue while new bone is deposited on and within the implant channel [4]. Therefore, these implants theoretically eliminate the need for a second operation, reduce the risk of infection, provide adequate strength during the critical early stages of healing, and permit a progressive transfer of stress to bone [4,5,6,7,8,9].
One promising implant material is polydioxanone. Initially developed as biodegradable surgical suture material, polydioxanone is now available as 1.3- and 2.0-mm-diameter pins. The pins have been used successfully for fixation of osteochondral fragments [10] and allografts [11,12,13] in the knee and hand and for stabilization of osteotomies and interphalangeal arthrodesis in the foot [7, 14,15,16,17,18,19,20].
Despite the clinical potential of polydioxanone implants, a resorption rate in humans has not been fully elucidated. In a recent study of rabbits, intraosseous polydioxanone pins were absorbed in 210 days [21]. To our knowledge, only three published studies, one using conventional radiography [14] and two using MR imaging [5, 22], have assessed polydioxanone pin resorption in human subjects. These studies indicate that pin channels may persist at least 3 and 6 months after surgery in the human knee and foot, respectively. In addition, although occasional cases of postoperative osteolytic and foreign-body reactions caused by polydioxanone pins [23] have been reported, the incidence and severity of adverse inflammatory reactions elicited by these devices have not been determined.
To our knowledge, a systematic investigation delineating the appearance, documenting the resorption, and determining the rate of inflammatory reactions of polydioxanone pins in human knees has not been performed. In this study, MR imaging was used to assess polydioxanone implants that were placed to secure osteochondral shell allografts [11,12,13, 24] in a series of knees studied at various postoperative time points. Emphasis was placed on graft fixation, pin-channel morphology, edema around pin channels, and cartilaginous defects at pin insertion sites.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Surgical Technique
Osteochondral allograft placement was performed by two experienced
orthopedic surgeons using the standard published technique
[11]. At surgery, the chondral
defect (Fig. 1A) is excised,
squared off, abraded down to bleeding subchondral bone, and measured. A 5- to
6-mm-thick osteochondral shell is removed from the orthotopic site of a
cadaveric donor knee, trimmed down to the precise size, and press-fitted into
the defect (Fig. 1B). If, in
the opinion of the surgeon, the graft is mechanically unstable in the host
bed, supplementary fixation is achieved with the insertion of 1.3-mm
biodegradable polydioxanone pins (Orthosorb absorbable pins; Johnson &
Johnson Professional, Raynham, MA). For each pin, a 1.3-mm-diameter channel is
drilled through the allograft and into the cancellous host bone until the
desired depth (up to 40 mm) is achieved, and the implant is then inserted into
the channel with the use of a proprietary applicator device (Johnson &
Johnson Professional). If the graft is mechanically stable with press-fitting
alone, supplementary fixation is not attempted.
|
|
As opposed to nonpatellar grafts in which only the defective portion of the articular surface is replaced, for patellar grafts, the entire patellar surface is removed even if only a portion is defective. Thus, unlike nonpatellar grafts, which are fitted snugly into precisely measured, squared-off surgical beds, patellar grafts are laid flush against a flat bed of host cancellous bone and are not buttressed along their margins by an osteochondral rim of host tissue. Therefore, supplementary fixation is mandatory to prevent slippage or dislodgment.
Two to four pins generally are used per graft, with a range of from one to seven. Among the 59 allografted surfaces that required supplementary pin fixation in this series, one graft had one pin, 20 grafts had two, 23 had three, 12 had four, two had five, and one had seven.
After surgery, patients are instructed to remain non-weight-bearing and undergo intensive physical therapy for a minimum of 12 weeks.
MR Imaging
Unenhanced MR examinations were performed on a 1.5-T unit (Signa; General
Electric Medical Systems, Milwaukee, WI) using 4-mm contiguous T1-weighted
spin-echo (TR/TE, 600/15), 4-mm contiguous proton density-weighted (3000/40),
T2-weighted fast spin-echo (3000/63), and 3.3-mm contiguous three-dimensional
spoiled gradient-recalled sequences (47/7; flip angle, 60°). All images
were acquired with fat saturation except the T1-weighted spin-echo sequences.
Matrix size was 256 x 256. Sagittal and coronal images (field of view,
12 cm) were obtained routinely, and additional axial images (field of view, 10
cm) were obtained using the same parameters for evaluation of pins in patellar
and trochlear grafts. Contiguous sequences were used to ensure inclusion of
all pin channels.
Scans were obtained 3, 6, 12, 24, or 36 months after surgery. The 175 pins in 59 grafts were imaged a total of 271 times (approximately 1.5 MR evaluations per pin). Eighty-six pins were imaged once (9 pins at 3 months, 15 at 6 months, 28 at 12 months, 4 at 24 months, 30 at 36 months). Eighty-two pins were imaged twice (13 pins at 3 and 6 months, 27 at 6 and 12 months, 17 at 12 and 24 months, and 25 at 24 and 36 months). Seven pins were imaged three times (all at 3, 6, and 12 months). The 13 grafts that contained no pins were imaged a total of 22 times. Four grafts were imaged once (1 at 3 months and 3 at 12 months), and nine grafts were imaged twice (4 at 3 and 6 months, 3 at 6 and 12 months, 1 at 12 and 24 months, and 1 at 24 and 36 months).
MR Interpretation
In consensus, two osteoradiologists who were unaware of clinical outcome,
elapsed time after surgery, the number and type of grafts placed, the time
interval after surgery, and the number of pins inserted evaluated each MR
imaging examination at each time point in the following way.
The number of visible grafts was recorded. The seating of each graft in the host bone was evaluated. Gaps between the graft marrow and the subjacent host marrow and other evidence of graft dislodgment were sought. For nonpatellar grafts, junctional clefts and stepoffs at graft margins were documented and measured. For patellar grafts, the alignment of graft and host margins was assessed.
The number of pin tracts visible in each graft on T1-weighted, T2-weighted, and spoiled gradient-recalled images was determined, and the signal intensity on each sequence was recorded. The length and width of each pin tract were measured to the nearest millimeter using mechanical calipers referenced against the scale on the image. Measurements of individual pins generally were identical on different MR sequences, although pin channels were, occasionally, obscured by surrounding edema on T2-weighted images, making measurements unreliable with this sequence in these instances. The imaging sequence in which the pins were most conspicuous was determined. The edges of pin tracts visible on T1-weighted images were scored subjectively as indistinct, intermediate, or sharp. T1-weighted sequences were used for this assessment because of their better spatial resolution compared with that of the other sequences.
Edema was defined as regions of high signal intensity within the host or graft marrow on T2-weighted images surrounding or directly abutting pin channels. Edema was considered focal if it was localized to only a portion of the pin; it was considered diffuse if it involved the entire or almost entire length of the pin. The number of pins associated with focal or diffuse edema was determined. High signal intensity within the graft or host marrow distinct from the pin was considered to represent nonspecific postoperative edema or an immune response to the graft but was not thought to reflect inflammation elicited by the pin.
The number and diameter in millimeters of fullthickness cartilage defects at pin entrance sites were documented.
Analysis
Overall, 271 MR imaging pin evaluations were performed. The imaging data at
each time point (3, 6, 12, 24, or 36 months) were pooled for analysis. Thus,
29 pins were analyzed at 3 months, 62 at 6 months, 79 at 12 months, 46 at 24
months, and 55 at 36 months. Five control grafts without pins were analyzed at
3 months, seven at 6 months, seven at 12 months, two at 24 months, and one at
36 months.
After the data for individual MR examinations were recorded, serial examinations were reviewed retrospectively. Images from these examinations were directly compared to assess subtle interval changes in graft position and to confirm that individual pin channels seen on successive studies represented the same channels rather than sampling error with different pins seen and missed at different times.
Statistical comparisons were performed using the unpaired Student's t test.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
The 59 grafts in which pins were used for supplementary fixation were evaluated a total of 92 times. Grafts always were correctly recognized within host tissue, usually because of obvious signal-intensity abnormalities in or adjacent to the graft. Rarely, grafts blended almost imperceptibly into the host bone and were identified on the basis of subtle morphologic differences. All grafts were seated at the operative site without evidence of dislodgment or loosening. For patellar grafts, the margins of transplanted and host tissue were aligned within 1-3 mm. The nonpatellar grafts fit snugly within the osteochondral beds in which they had been secured. Junctional clefts and stepoffs at graft margins were usually immeasurably thin and were rarely as large as 1-2 mm. The 30 grafts that were studied serially were stable in position between examinations.
As opposed to grafts, which invariably were recognizable, pin channels were not always visible, and the percentage of visible pins decreased as a function of time on all imaging sequences (Table 1). More than 80% of the pins were visible on both T1- and T2-weighted images at 3-6 months after surgery, whereas no more than 20% of the pins were visible on these sequences at 24-36 months after surgery (Table 1). The slight increase in the percentage of visible pins between 24 and 36 months reflects the fact that the same pins were not necessarily evaluated at each time point. In the early postoperative period, pins were identified most easily on T1-weighted images. At and after 12 months, the pins generally were most conspicuous on T2-weighted images. The number of visible pin channels in each graft never exceeded the number of pins actually placed.
When visible, pins were seen as straight linear regions of abnormal signal intensity that were up to 38 mm long. No pin channels were angulated or discontinuous. With only three exceptions, channels were of low signal intensity on T1-weighted images at all time points (Table 2 and Fig. 2A,2B). During the first 6 postoperative months, channels could be of either low or high signal intensity on T2-weighted images but invariably were of high signal intensity afterward (Table 2 and Fig. 3A,3B). In five of the pins studied serially, the signal intensity changed from low to high on successive examinations (Fig. 3A,3B).
|
|
|
|
Of the 89 pins that were studied serially, 30 pins on T1-weighted images, 31 pins on T2-weighted images, and 32 pins on gradient-recalled images were visible on the initial study but not on the next, presumably having been resorbed entirely during the interval (Fig. 2A,2B). Fifty-nine pin channels were revisualized on at least one imaging sequence on follow-up studies in precisely the same location relative to adjacent landmarks. Because these channels were situated identically on each MR examination, the findings almost certainly represented the same channels rather than sampling differences. Of the 59, 11 narrowed by 1-2 mm on successive scans but did not disappear (Fig. 2A,2B). One pin channel widened slightly (from 1 to 2 mm) between 3 and 6 months, but it became more indistinct in the process, and it then disappeared by 12 months. There was no instance in which a pin channel that had not been apparent on an earlier scan became visible on a subsequent scan.
Although individual pins tended to disappear or become narrower, the remaining pins were essentially constant in size, with mean lengths measuring between 23 and 28 mm and mean widths measuring between 1.2 and 1.6 mm across the various time points (Tables 3 and 4). Despite the constant size, pins tended to become progressively more indistinct (Tables 3 and 4). Among visible pins, the mean thickness of pins in patellar grafts (2.0 ± 0.3 mm) was greater (p < 0.01) than that of pins in nonpatellar grafts (1.4 ± 0.1 mm). The three thickest pins (2 with a 4-mm width and 1 with a 6-mm width) were patellar (Fig. 4).
|
Overall, nearly one fifth of pin channels were associated with edema (Table 5). In approximately two thirds of the cases, the edema was focal and in one third it was diffuse. Subjectively, the edema was relatively minor, ranging from 2 to 7 mm in width. The prevalence of pin-related edema was highest at 3 postoperative months (almost 40% of the pins) and was stable at between 11% and 16% for the later postoperative time points.
In pins studied serially, edema generally remained stable or diminished with time. In 10 instances, edema resolved (Fig. 5A,5B) (generally at 3-6 months, but on one occasion between 24 and 36 months). There were only two cases in which edema developed from one study to the next: one in a patient examined at 6 and 12 months and the other in a patient examined at 12 and 24 months.
|
|
Focal cartilage defects measuring 1-2 mm in diameter were identified at insertion sites in 13% of the pins (Fig. 6). These defects were more common in the early postoperative time periods (approximately one third of pins at 3-6 months) rather than in the later postoperative time periods (<5% of pins at and after 12 months) (Table 5). In five pins studied serially, defects resolved between successive examinations.
|
The 13 control grafts were imaged a total of 22 times. All were appropriately seated within the host osteochondral bed. Nine of the grafts were examined serially and were stable in position between examinations. In no instances were pin channels, pin-related edema, or cartilage defects falsely identified in these grafts.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
All these implants degrade within host tissues by hydrolysis and nonspecific enzymatic activity [9, 21, 23, 29]. In vivo, the resultant synthetic debris is cleared predominantly by tissue macrophages [9]. Of the three implant types, polyglycolic acid implants lose strength and are absorbed most rapidly [4, 16, 26]. This feature is clinically important because the resorption rate constitutes a critical factor for the acceptance of these devices by the host [9]. In humans, polyglycolic acid reportedly is absorbed in 3 months [26]. This degradation rate may exceed the capacity of tissue macrophages and local blood flow to absorb and transport the breakdown products away from the operative site. Polymeric liquid debris may accumulate, leading to latent foreign-body reactions with osteolytic changes [15, 29, 33] and draining sinus tracts requiring surgical débridement [14, 15, 25, 29, 30]. Polylactic acid implants have the longest degradation time, with an estimated duration of several years [26, 27, 31], and probably cause osteolytic and foreign-body reactions with the lowest frequency. Their protracted resorption, however, delays osseous ingrowth into the implant channel and may compromise the structural competence of the host bone [26]. Because of their intermediate resorption, polydioxanone implants offer a potential compromise between an acceptably low incidence and severity of inflammatory responses and an acceptably rapid restoration of normal bony integrity [17,18,19, 23, 26].
Despite the importance of this parameter, the precise resorption rate of polydioxanone implants in humans is unknown. Atkinson et al. [21] have shown that polydioxanone pins placed within the rabbit knee are completely resorbed in 210 days, but the resorption rate in humans may be much slower owing to differences in basal metabolic rates between laboratory animals and humans [29]. Only three published reports have addressed this issue in human subjects. In a retrospective review of 12 patients who underwent metatarsal osteotomies, Friend et al. [14] observed persistent radiographic lucency of pin tracts at least 6 months after surgery. In five patients, moreover, the diameter of the lucency was greater than that of the pin itself. In another study, a case report of a patient with polydioxanone pin fixation for osteotomy in the foot, Bashara et al. [5] observed abnormal signal intensity within a pin tract on MR imaging at 26 postoperative weeks. Bone had not yet grown into and replaced the tunnel. Similarly, in a case report of a patient with polydioxanone pin fixation of osteochondritis dissecans of the knee, Smith et al. [22] observed linear decreased signal intensity in pin channels at 3 postoperative months.
This study systematically investigated a relatively large cohort of patients with polydioxanone pin fixation of osteochondral fragments of the knee. Because of its superior delineation of soft tissue, cartilage, and bone marrow, MR imaging was used for evaluation of pin resorption and detection of inflammatory tissue reactions. This study establishes that pin channels can be reliably depicted on MR imaging. More than 80% of pin channels were visible at 3-6 months after surgery, a time period before the expected complete resorption of these implants. During the early postoperative period, pin channels were seen best on T1-weighted images, probably because of the greater spatial resolution of T1-weighted sequences and because many pins were of low signal intensity on T2-weighted images. At later time points, pin channels developed high signal intensity on T2-weighted sequences, and pins generally were most conspicuous with T2-weighting. Pin channels were never falsely identified on unsecured transplanted surfaces, and the number and length of identified pin channels never exceeded the number or length of pins actually placed.
This study confirms that polydioxanone pins resorb in human knees. Approximately 80% of pins could not be seen on any imaging sequence at and after 2 years, presumably having resorbed and been replaced by bone. The change from low to high signal intensity on T2-weighted images in many pins after 3 postoperative months may represent the chemical hydrolysis documented in animal studies. Thus, low signal intensity at 3 months may reflect actual polydioxanone material, whereas high signal intensity may represent hydrolyzed debris or fluid filling empty pin channels. Although pins did not change in signal intensity on T1-weighted images, the pins tended to become more indistinct, to become thinner, or to disappear entirely. The mean width of remaining pins at any given time point was remarkably constant; this finding makes sense because thinner pins are more likely to resorb and disappear between studies, whereas wider pins may become narrower but not disappear, resulting in a stable mean width.
All pin channels were 1.3 mm in diameter at the time of surgery but many channels were 2 mm or more in diameter at the time of imaging. On occasion, the interval widening that was noted on MR imaging may have been artifactual and related to sampling, volume averaging, limitations in spatial resolution, or a combination of these factors. Usually, however, the widening was probably real, especially for pin channels measuring at least 3 mm in diameter, and likely represented postoperative osteolysis. Pin channels in patellar grafts were significantly wider than those in nonpatellar grafts, suggesting greater osteolysis. This observation may reflect the greater mechanical stress imposed on patellar pins, which provide the only source of stability in patellar grafts until osseous fusion, whereas marginal buttressing contributes to the mechanical stability of nonpatellar grafts. On occasion, however, pin channel widening may have represented a healing response rather than osteolysis. For example, the single pin that became wider between successive examinations also became more indistinct during the same time interval and later disappeared.
Edema, defined as regions of high signal intensity on T2-weighted images spatially related to pin channels, presumably represented nonspecific foreign-body inflammation to synthetic debris, as documented in animal studies [21]. Trauma related to pin insertion may have contributed to edema in some patients, especially during the early postoperative period. Edema was relatively common during the first 3 months and relatively uncommon afterward. Subjectively, edema was usually relatively minor, measuring only a few millimeters in extent. In approximately one third of the cases in which edema was seen, edema involved the pin channel diffusely, whereas in two thirds of the cases, edema was focal. In pins studied serially, edema tended to decrease between successive examinations. In two of the patients examined serially, however, edema developed between scans, suggesting that residual pin material can incite inflammation even after the early postoperative period.
Focal chondral defects at pin insertion sites were common during the early postoperative period but disappeared later, implying that polydioxanone material resorbs within the cartilaginous and the osseous portions of the pin channel and that the cartilaginous channel heals.
In this study, polydioxanone pins were effective in securing shell osteochondral allografts. All grafts in which supplementary pin fixation was used were appropriately seated within the host bed at initial imaging. Grafts studied serially remained in stable position. Because channels were uniformly straight without angulation or discontinuity, there was no evidence of pin fracture.
A few limitations of this study merit discussion. Because routine histologic evaluation of our patients was obviously not possible, we could not confirm the significance and interpretation of MR findings. Because of limited spatial resolution, pin diameters were measured to the nearest millimeter on MR examinations even though the channels were drilled precisely to 1.3 mm at surgery. At any given time point, all grafts and all pins were pooled together for analysis, although there were different numbers, types, and sizes of grafts and different numbers of pins per graft and pins per knee. Therefore, not all grafts and pins were subject to the same local biomechanical and biologic environment. Moreover, not all grafts and pins were imaged the same number of times or at the same time points. As a result, some grafts and some pins were relatively overrepresented, whereas some were relatively underrepresented in the pooled analysis. Finally, this study investigated the use of polydioxanone pins in a select group of patients who had received osteochondral allografts, an experimental procedure requiring great technical expertise, in a tertiary referral center. On a national level, biodegradable pins likely would be used more commonly in the fixation of lesions associated with osteochondritis dissecans and osteochondral fractures [10, 28], but for logistic reasons, patients with these abnormalities were not recruited. It is conceivable, although unlikely, that immune-mediated reactions against the transplanted graft affected the responses to the pins documented in this investigation.
Despite these limitations, this study establishes a spectrum of MR appearances for pin channels at different postoperative time points. This study suggests successful, progressive resorption in most pins with relatively infrequent, minor inflammatory responses. Focal cartilage defects at the insertion sites are common early postoperatively but tend to resolve. Awareness of these appearances is beneficial because as the use of biodegradable pins increases, these devices will be imaged with greater frequency. In our study, T1- and T2-weighted sequences generally were sufficient for evaluation of pins, but depending on institutional and personal preferences, proton density-weighted and gradient-recalled sequences may be desired for assessment of the rest of the knee.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
