(Radiology. 2001;218:278-282.)
© RSNA, 2001
Osteochondral Lesions of the Talus: Computer-assisted Retrograde Drilling—Feasibility and Accuracy in Initial Experiences1
Reto J. Bale, MD,
Christian Hoser, MD,
Ralf Rosenberger, MD,
Michael Rieger, MD,,
Karl P. Benedetto, MD and
Christian Fink, MD
1 From the Interdisciplinary Stereotactic Interventional Planning Laboratory at the Department of Radiology (R.J.B., M.R.), the Department of Traumatology (C.H., R.R., K.P.B., C.F.), and Institute of Anatomy (R.R.), University of Innsbruck, Anichstrasse 35, A-6020 Innsbruck, Austria. From the 1999 RSNA scientific assembly. Received December 6, 1999; revision requested January 10, 2000; final revision received April 25; accepted May 8. Supported by the Lorenz Boehler Gesellschaft-Verein zur Förderung der Forschung auf dem Gebiet der Unfallchirurgie (Project 2/99). Address correspondence to R.J.B. (e-mail: reto.bale@uibk.ac.at).
 |
ABSTRACT
|
|---|
The authors developed a minimally invasive method with computer-assisted navigation for retrograde drilling of osteochondral lesions of the talus. Planning of the pathway and adjustment of the targeting device were performed outside the operating room. In 10 cadavers and four patients, accuracy of pin placement was in the range of 1.0–3.5 mm.
Index terms: Computed tomography (CT), three-dimensional, 4641.12117 • Computers, examination control • Osteochondritis dissecans, 4641.442 • Stereotaxis, 4641.1267
|
INTRODUCTION
|
|---|
For a preoperative surgical plan, it is important to distinguish between stable and unstable lesions in talar osteochondritic lesions (1). Loose fragments have to be removed in most cases, and débridement of the defect is performed. The surgical treatment advocated for stage I–III lesions (2) is drilling through the sclerotic margin of the lesion (1,3,4–7) to improve perfusion.
Intraoperative visualization can be achieved with open dissection with or without osteotomy of the malleolus, arthroscopy, or fluoroscopy. A high percentage of osteochondral lesions are located in the posterior aspect of the talar dome; thus, they are difficult to visualize and access through anterior arthroscopy portals. Transmalleolar (7) and anteromedial (8) approaches, as well as drill guides (5,6), have been described to help access the posteromedial part of the talus. To leave the cartilage surface intact, drilling has to be performed in a retrograde fashion from the talar body into the dome (6).
Our goal was to develop a minimally invasive method for accurate retrograde drilling of osteochondral lesions of the talar dome with use of computer-assisted navigation. The purpose of this study was to evaluate an immobilization-fixation technique that was developed for convenient noninvasive immobilization and external registration (Fig 1). The accuracy of drilling into well-defined landmarks in the medial talar dome was evaluated in a cadaver study and in four patients.
|
Materials and Methods
|
|---|
Fixation Technique for Lower Leg
We fabricated a fixation device of the leg. The ankle joint was positioned in 10° of plantar flexion regardless of the location of the lesion. In the first step, a plastic foil (Medical Intelligence, Schwabmùnchen, Germany) was wrapped around the leg from the toes to the region below the knee to avoid contact of the cast layers with the skin. Second, an elastic fiberglass stocking (STS Copy-Sock; Goetz, Göppingen, Germany) was rolled onto the leg in a tight fit. After 5 minutes of hardening, it was taken off by making a longitudinal cut anteriorly. This gave an accurate mold. Six layers of synthetic cast material (Scotchcast; 3M Health Care, St Paul, Minn) were used to produce a rigid construct, and a Plexiglas connector was integrated into the cast. The connector was positioned with respect to the anticipated surgical approach for drilling so that it did not obstruct the entry area. This allowed reproducible rigid attachment of the cast to a base plate. Finally, the cast was cut open anteriorly and posteriorly to produce a medial and lateral shell that was held together with hook and loop straps during imaging and the surgical procedure. Between six and 10 radiopaque markers (Philips Medical Systems, Best, the Netherlands) were glued to the cast in an even distribution from 10 cm proximal to 10 cm distal to the ankle.
Imaging
Transverse computed tomography (CT) (HiSpeed CT/i Advantage; GE Medical Systems, Milwaukee, Wis) was performed (120 kV, 120 mAs, section thickness of 1 mm, pitch of 2, increment of 1 mm, 512 x 512-cm matrix, 24-cm field of view, and 0° gantry tilt) with the leg immobilized in the cast. The data set was transferred to the navigation system located in the planning laboratory by means of a local network.
Navigation
Navigation systems.—Navigation systems are based on optical localization technology with use of a powerful workstation and a system to measure the optical position. The system used in this study (Philips Medical Systems) comprised a camera array with two two-dimensional cameras (infrared-sensitive charge-coupled device), a variety of handheld pointer instruments, and a dynamic reference frame. The optical sensor array tracked the spatial coordinates of light-emitting diodes attached to the instruments. The workstation calculated the coordinates of the sensor and showed the actual position of the instrument in relation to the preoperative data set. We used standard commercially available stereotactic navigation systems (EasyGuide, Philips Medical Systems; StealthStation, Sofamor Danek, Memphis, Tenn; Surgical Tool Navigator, Zeiss, Jena, Germany). The software and hardware capabilities, including the accuracy, were comparable for each and in comparison with those of alternate commercially available optical localizing systems, but each system had its own characteristics. All three systems implemented a guidance software module that allowed both development of a surgical plan and interactive execution.
Definition of path.—Three-dimensional reconstruction images of the synthetic cast material, skin, bone, and radiopaque markers were derived from the original CT data set (Fig 2). The pathway was determined by the surgeon with the three-dimensional navigation system in multiplanar sections, including transverse, coronal, and sagittal views. In addition, the reformatted planes along the defined path were visualized. In all patients, the lesions were located in the medial talar dome. In the cadavers, well-defined individual landmarks located at the medial aspect of the talar dome, such as diminutive cysts or sclerotic areas, served as target points. We planned a drill path that started at the lateral process of the talus and led into the lesion on the medial side of the talar dome.
View larger version (195K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2. CT images were obtained with the three-dimensional navigation system to determine the target point of osteochondritis dissecans at the medial aspect of the talar dome. Target point (crosshair in A-D, arrowheads in A-C) is selected on reconstructed two-dimensional sagittal (A), coronal (B), and transverse (C) images and on the three-dimensional image (D). Projection of path (arrows) is visible on A-C.
|
|
Registration.—In the absence of the patient (or the cadaver), the fixation device was repositioned on the base plate by means of the Plexiglas connector and straps (Fig 3). Next, the registration procedure was performed, in which the image of the leg immobilized in the cast was related to the real fixation device. Image-to–fixation device registration was performed by defining the virtual fiducial markers on the imaged data set and indicating the respective real fiducial markers on the synthetic cast material by using the pointer of the navigation system. The system registered the image and marker positions and derived the root-mean-square error, which is an indicator of registration accuracy. The root-mean-square error was accepted only if it was less than 0.8 mm. If it was higher, the registration procedure was repeated. By using the pointer of the navigation system, the preplanned entrance point on the cast was determined and marked. A 5-cm-diameter circle was cut out of the cast to serve as the entrance for the drill.
Targeting.—The central element of the patented targeting device (9) (EasyTaxis; Philips Medical Systems) is a spheric alignment body (trapped ball) that rotates freely in a bearing and can be locked by using a screw (Fig 4). The alignment body contains a large central cylindrical hole to accept the probe or different surgical instruments (eg, endoscope, needles, pins) and is mounted to the distal end of a mechanical arm with 6 df.
View larger version (61K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4. Schematic of laboratory setup. The targeting device is adjusted with respect to the cast (1) with use of the navigation system. Operator has introduced the pointer into the aiming device. Movements of the pointer (2) are detected by the camera (3) and depicted on the monitor (4). Distal view (Insert 1) shows a Plexiglas connector between the fixation device and base plate. Transparent illustration (Insert 2) shows the targeting device aimed at the entry (E, arrow) and target (T, arrowhead) in the talus.
|
|
The mechanical arm is fastened to the base plate, which then holds both the immobilization device (ie, cast) and the mechanical arm in a rigid and reproducible fashion. The pointer of the navigation system is positioned in the cylindrical hole in such a manner that the tip of the pointer is in the geometric center of the trapped ball. With use of the guidance software of the navigation system, actual targeting was performed in a two-step procedure. In step 1, the tip of the pointer (which is locked in the geometric center of the trapped ball) was arranged on the line that extends through the predefined entrance position and the target position (ie, alignment line). The mechanical arm was fastened. In step 2, the pointer was rotated in the ball joint until the longitudinal axis of the probe was aligned along the preplanned path (ie, alignment line). Then the ball joint was fastened.
The depth of the pin insertion was calculated by the navigation system. The probe was removed, and the adjusted targeting device was kept in position until the surgical procedure was performed. In the patient study, the fixation device with the connector, the base plate, the straps, and the targeting device were sterilized with formaldehyde gas.
Surgical Procedure
After routine sterile preparation and draping and before the induction of general anesthesia, the patient’s leg was repositioned in the sterile fixation device (Fig 5). After we had confirmed a perfect fit of the cast, the patient received general anesthesia. The alignment device was then repositioned. A 2.4-mm-thick pin with the precalculated length was mounted in a Jacob chuck of the drill, and was then advanced through the preset aiming device. Then the cast was removed from the leg. The position of the pin was documented with fluoroscopy, and a parallel drill guide was used to advance additional pins into larger lesions if desired. In all patients, arthroscopy was then performed to remove loose fragments and to check the integrity of the cartilage surface.
Evaluation of Accuracy
Patient studies.—Four consecutive patients (two men and two women, aged 18–35 years [mean age, 23 years]) with osteochondritis dissecans of the talus were treated with this method. The lesion was classified as stage II in one patient, stage III in two, and stage IV in one according to the system of Berndt and Harty (2). All lesions were located at the medial aspect of the talar dome. Informed consent was obtained from all patients after the nature of the procedure had been fully explained. The study was approved by our institutional review board. In all patients, the pin position was visualized with intraoperative fluoroscopy, and CT (Fig 6) was then performed to help identify the drill hole with respect to the osteochondral lesion.
View larger version (183K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6. Patient postoperative CT scan (reconstructed section through the three drill holes, near the coronal plane) was acquired to confirm the drill hole positions. Pins were removed at the end of the procedure. The middle hole (arrow) was planned and guided; the additional two holes were made with parallel drill guides.
|
|
Cadaveric studies.—In 10 cadaveric specimens, the position of the pin tip was evaluated by means of anatomic dissection.
Procedure duration.—The duration of the different tasks involved in the procedure was evaluated.
|
Results
|
|---|
The fixation device was fabricated immediately after the decision was made to perform computer-assisted drilling and lasted 30 minutes. CT was performed as an outpatient procedure. The patient was admitted to the hospital on the morning of surgery. The repositioning procedure and the drilling procedure lasted 5–10 minutes. The conventional arthroscopic procedure was not different from a routine arthroscopic intervention in the ankle joint. None of the patients reported pain or discomfort related to the immobilization method. The time for path planning and adjustment of the targeting device with the guiding software was between 20 and 30 minutes.
In the 10 cadavers and four patients, the approximate side deviation of the pin tip position determined at cadaveric dissection and CT, respectively, versus that in the plan established with the navigation system was in the range of 1.0–3.5 mm. Accuracy was slightly worse in the patients (2.5, 3.0, 3.0, and 3.5 mm) than in the cadavers (mean, 2.05 mm; SD, 0.96 mm). We saw neither pins that were advanced too far nor pins that would not reach the subchondral area. In all four patients, the lesion was reached, and the pin perforated the sclerotic margin of the lesion. As confirmed with arthroscopy, the pin did not perforate the cartilage unintentionally in any case. Surgery time with this technique compared with the conventional technique at our institution was reduced by 30 minutes.
|
Discussion
|
|---|
Fluoroscopy is the imaging method of choice to perform a variety of surgical procedures such as fracture reduction, osteotomy, and bone tumor biopsy. Thus, it seems to be the ideal tool to help localize osteochondral lesions before retrograde drilling. Although inexpensive and readily available, fluoroscopy has several important limitations (10). Fluoroscopic images are two-dimensional, and the surgeon must mentally reconstruct the three-dimensional situation. Continuous use of the fluoroscope is often necessary during surgery, which results in a cumulative radiation exposure to the surgeon and the patient (11). Lesions with late stages of osteochondritis dissecans cannot be detected well with fluoroscopy. In addition, accurate targeting of a lesion is technically challenging at the first attempt, which often necessitates multiple drillings. In stage III and IV lesions with an antegrade defect, drilling guided with arthroscopy is possible for anterior locations but increasingly difficult with further posterior position. The use of drill guides (6) is technically insufficient if the lesion cannot be seen at arthroscopy. If the surgeon wishes to drill into the lesion from more than one direction, a second or third targeting device could be used for additional drilling procedures. With this procedure, the duration of surgery to target a lesion could be markedly reduced, and exposure of the surgeon and patient to the image intensifier could be nearly avoided.
Frameless stereotactic navigation systems allow visualization of the actual location of the probe on the patient’s preoperative images as measured with a three-dimensional digitizer (12). A prerequisite for successful application of frameless stereotactic navigation systems is the registration procedure, in which the patient’s image data set is related to the real patient.
At our institution, navigation systems in combination with aiming devices have already been used successfully for brain tumor biopsy, fractionated computer-assisted interstitial brachytherapy (13), and trigeminal ganglion puncture. All of these procedures were performed in the head region, however, which can be immobilized with dental cast–based systems (13,14) or with stereotactic frames (15). In computer-assisted total hip replacement and total knee arthroplasty, invasive pins are used to immobilize and register the patient’s body (10,16,17). Preoperative placement of these artificial fiducial markers has certain drawbacks: an extra surgery is needed, and there is a risk of infection. In addition, patients experience postoperative pain (17).
To overcome inconvenient invasive registration and immobilization procedures, we adapted noninvasive immobilization methods to our purposes. The independence in time and location of immobilization with synthetic cast material, image acquisition, path planning, adjustment of the targeting device, and treatment was appreciated by both the patients and the medical personnel.
This approach is time-consuming in the preoperative phase. Although the improvements with our method resulted in a reduction in surgical time, the whole procedure, including immobilization with the synthetic cast material, planning, and targeting, requires an additional 50–70 minutes. The time factor is affected by a learning curve for the medical personnel as the regimen becomes routine.
Arthroscopy was performed in the patients for two reasons. First, the integrity of the cartilaginous surface could be investigated, and second, the joint could be inspected for loose osteochondral fragments. With improved technology for magnetic resonance arthrography, which would allow better visualization of the chondral surface, conventional arthroscopic inspection may be obviated in selected cases. This would result in a considerable reduction in treatment time. For the majority of patients, however, conventional arthroscopy will remain an essential part of the procedure.
In conclusion, our experience suggests that frameless stereotactic navigation systems in combination with rigid reproducible fixation allows preplanned retrograde drilling of osteochondral lesions in the talus with accuracy that is sufficient to reduce surgery time and invasiveness. Computer-assisted navigation allows interactive target selection on preoperative CT images without the restrictions of surgical access. With this method, image acquisition, adjustment of the targeting device, and surgery can be dissociated from each other.
|
FOOTNOTES
|
|---|
Author contributions: Guarantors of integrity of entire study, C.F., K.P.B.; study concepts, all authors; study design, all authors; definition of intellectual content, R.J.B., C.H., C.F.; literature research, R.J.B., R.R., C.H., C.F.; clinical studies, all authors; experimental studies, all authors; data acquisition, R.J.B., C.H., C.F., R.R.; data analysis, C.H., R.R.; manuscript preparation, R.J.B., C.H., C.F.; manuscript editing, R.J.B., C.H.; manuscript review, R.J.B., C.H., C.F., K.P.B.
|
REFERENCES
|
|---|
-
Clanton TO, DeLee JC. Osteochondritis dissecans: history, pathophysiology and current treatment concepts. Clin Orthop 1982; 167:50-64.
-
Berndt AL, Harty M. Transchondral fractures (osteochondritis dissecans) of the talus. J Bone Joint Surg Am 1959; 41:988-1020.[Abstract/Free Full Text]
-
Ferkel D, Karzel RP, DiPaola J, Del Pizzo W, Friedmann MJ, Fischer SP. Arthroscopic treatment of anterolateral impingement of the ankle. Am J Sports Med 1991; 19:440-446.[Abstract/Free Full Text]
-
Baker CL, Andrews JR, Ryan JB. Arthroscopic treatment of transchondral talar dome fractures. Arthroscopy 1986; 2:82-87.[Medline]
-
Bryant DD, III, Siegel MG. Osteochondritis dissecans of the talus: a new technique for arthroscopic drilling. Arthroscopy 1993; 9:238-241.[Medline]
-
Taranow WS, Bisignani GA, Towers JD, Conti SF. Retrograde drilling of osteochondral lesions of the medial talar dome. Foot Ankle Int 1999; 20:474-480.[Medline]
-
Kumai T, Takakura Y, Higashiyama I, Tamai S. Arthroscopic drilling for the treatment of osteochondral lesions of the talus. J Bone Joint Surg Am 1999; 81:1229-1235.[Abstract/Free Full Text]
-
Flick AB, Gould N. Osteochondritis dissecans of the talus (transchondral fractures of the talus): review of the literature and new surgical approach for medial dome lesions. Foot Ankle 1985; 5:165-185.[Medline]
-
Dorward NL, Alberti O, Dijkstra A, Buurman J, Kitchen ND, Thomas DG. Clinical introduction of an adjustable rigid instrument holder for frameless stereotactic interventions. Comput Aided Surg 1997; 2:180-185.[Medline]
-
Joskowicz L, Milgrom C, Simkin A, Tockus L, Yaniv Z. FRACAS: a system for computer-aided image-guided long bone fracture surgery. Comput Aided Surg 1998; 3:271-288.[Medline]
-
Sanders R, Koval KJ, DiPasquale T, Schmelling G, Stenzler S, Ross E. Exposure of the orthopaedic surgeon to radiation. J Bone Joint Surg Am 1993; 75:326-330.[Abstract/Free Full Text]
-
Zinreich SJ, Tebo SA, Long DM, et al. Frameless stereotaxic integration of CT imaging data: accuracy and initial applications. Radiology 1993; 188:735-742.[Abstract/Free Full Text]
-
Bale RJ, Freysinger W, Gunkel AR, et al. Frameless stereotactic interstitial brachytherapy of head and neck tumors: initial experience. Radiology 2000; 214:591-595.[Abstract/Free Full Text]
-
Martin A, Bale RJ, Vogele M, Gunkel AR, Thumfart WF, Freysinger W. Vogele-Bale-Hohner mouthpiece: registration device for frameless stereotactic surgery. Radiology 1998; 208:261-265.[Free Full Text]
-
Maciunas RJ, Galloway RL, Jr, Latimer JW. The application accuracy of stereotactic frames. Neurosurgery 1994; 35:682-694.[Medline]
-
Kazanzides P, Mittelstadt B, Musits B, et al. An integrated system for cementless hip replacement: robotics and medical image enhance precision surgery. IEEE Eng Med Biol 1995; 12:307-313.
-
Kienzle T, Stulberg S, Peshkin M, et al. Total knee replacement: computer assisted surgical system using a calibrated robot. IEEE Eng Med Biol 1995; 12:301-306.
This article has been cited by other articles:
|
|
|
|
 
C. F. Hyer, G. C. Berlet, T. M. Philbin, and T. H. Lee
Retrograde Drilling of Osteochondral Lesions of the Talus
Foot & Ankle Specialist,
August 1, 2008;
1(4):
207 - 209.
[Abstract]
[PDF]
|
|
|
|
|
|
|
M. Easley, B. Chuckpaiwong, N. Cooperman, R. Schuh, T. Ogut, I. L.D. Le, and J. Reach
Computer-Assisted Surgery for Subtalar Arthrodesis. A Study in Cadavers
J. Bone Joint Surg. Am.,
August 1, 2008;
90(8):
1628 - 1636.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
M. Citak, D. Kendoff, M. Kfuri Jr, A. Pearle, C. Krettek, and T. Hufner
Accuracy analysis of Iso-C3D versus fluoroscopy-based navigated retrograde drilling of osteochondral lesions: A PILOT STUDY
J Bone Joint Surg Br,
March 1, 2007;
89-B(3):
323 - 326.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
M. Kono, M. Takao, K. Naito, Y. Uchio, and M. Ochi
Retrograde Drilling for Osteochondral Lesions of the Talar Dome
Am. J. Sports Med.,
September 1, 2006;
34(9):
1450 - 1456.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
M. J. Botte, K. A. Ezzet, L. L. Pacelli, M. J. Guzman, R. S. Meyer, M. J. Meunier, D. D. D'Lima, and C. W. Colwell
What's New in Orthopaedic Rehabilitation
J. Bone Joint Surg. Am.,
December 9, 2002;
84(12):
2312 - 2320.
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
