Posterior Ankle Arthroscopy AN ANATOMIC STUDY

Transcription

1 763 COPYRIGHT 2002 BY THE JOURNAL OF BONE AND JOINT SURGERY, INCORPORATED Posterior Ankle Arthroscopy AN ANATOMIC STUDY BY DAVID F. SITLER, MD, ANNUNZIATO AMENDOLA, MD, CHRISTOPHER S. BAILEY, MD, LISA M.F. THAIN, MD, AND ALISON SPOUGE, MD Investigation performed at Fowler Kennedy Sport Medicine Clinic, 3M Centre, University of Western Ontario, London, Ontario, Canada Background: Ankle arthroscopy has generally been performed with use of anterior portals with the patient in the supine position. Little has been published on ankle arthroscopy performed with use of posterior portals, particularly with the patient in the prone position. The purpose of the present study was to evaluate the relative safety and efficacy of ankle arthroscopy with use of posterior portals with the limb in the prone position. Methods: Thirteen fresh-frozen cadaver specimens were used. Posterolateral and posteromedial portals were established. Arthroscopy was performed, and the extent of the talar dome that could be visualized was marked. Four-millimeter plastic cannulae were filled with oil and were placed in the portals for use as reference landmarks on magnetic resonance imaging studies. The proximity of the portal cannulae to the adjacent structures was measured on standard magnetic resonance images and then during careful dissection. The distances measured by dissection were compared with the measurements made on magnetic resonance images. Results: An average of 54% (range, 42% to 73%) of the talar dome could be visualized. The average distance between a cannula and adjacent anatomic structures after dissection was 3.2 mm (range, 0 to 8.9 mm) to the sural nerve, 4.8 mm (range, 0 to 11.0 mm) to the small saphenous vein, 6.4 mm (range, 0 to 16.2 mm) to the tibial nerve, 9.6 mm (range, 2.4 to 20.1 mm) to the posterior tibial artery, 17 mm (range, 19 to 31 mm) to the medial calcaneal nerve, and 2.7 mm (range, 0 to 11.2 mm) to the flexor hallucis longus tendon. The magnetic resonance images demonstrated very similar distances except in the case of the distance between the posteromedial cannula and the tibial nerve, which often was difficult to specifically identify on magnetic resonance imaging studies. Conclusions: The findings of the present cadaveric study suggest that, with the patient in the prone position, arthroscopic equipment may be introduced into the posterior aspect of the ankle without gross injury to the posterior neurovascular structures. Limited clinical trials should be carried out to confirm this finding. A video supplement to this article is available from the Video Journal of Orthopaedics. A video clip is available at the JBJS web site, The Video Journal of Orthopaedics can be contacted at (805) , web site: A commentary is available with the electronic versions of this article, on our web site ( and on our quarterly CD-ROM (call our subscription department, at , to order the CD-ROM). In 1931, Burman concluded, following a cadaveric study, that the ankle joint was not suited for arthroscopy because the joint space was too narrow and the dome of the talus was too convex for adequate viewing 1. Improvements in technique, instrumentation, and distraction have established ankle arthroscopy as a useful and practical procedure. The expanding use of ankle arthroscopy and its risks have been well documented Reported complication rates for this procedure have ranged from 0% to 25% 2-7,11,12. Neurologic injuries, which ranged from 0.04% to 4.8%, accounted for the majority of the important complications in these series 2,3,5. These complications were primarily associated with anterior portals, with the patient in a supine position. It is sometimes difficult to access lesions in the posterior aspect of the ankle through anterior portals 13, and posterolateral, posteromedial, and trans-achilles tendon portals have been described in the literature 4,15. The safety, efficacy, and anatomic relationships associated with placement of the posterior portals have been investigated in only a few anatomic studies 13,15,16. Authors have generally discouraged the use of the posteromedial portal, citing the proximity of and high risk of injury to the tibial nerve and the medial calcaneal nerve as reasons 2,4,8, Most investigators have commented that the anterior portals and the posterolateral portal are safe and have recommended the use of those portals on the basis of studies in which patients were placed in the standard supine position 4,15-17,19. However, in our opinion, when posterior portals are used with the patient in the supine position, the procedure is difficult and there is a tendency to drift anteriorly as the portals are placed, which increases the risk of damage to neurovascular structures. Little has been published on ankle arthroscopy with the patient in the prone position. Zimmer and Ferkel discussed the use of posterior portals with the patient in the prone position but for endoscopy of the retrocalcaneal bursa only 20. We hypothesized that, with the ankle and foot in the prone position, posterior portal placement is safer

2 764 and allows better visualization of the posterior structures about the ankle. The purpose of the present study was to evaluate the safety and efficacy of the use of posteromedial and posterolateral portals for ankle arthroscopy with the foot in the prone position. We also wanted to evaluate the accuracy of magnetic resonance imaging as a tool for determining anatomic relationships in a cadaver study as its use may eliminate the distortion that occurs during dissection performed to identify the neurovascular structures adjacent to the arthroscopic cannulae. Materials and Methods hirteen fresh-frozen cadaver specimens that had been ob- from below-the-knee amputations were used for Ttained this study. The proximal part of the tibia was secured by a specimen holder (Pacific Research Laboratories, Vachon, Washington), with the foot in a prone position. Posterior ankle arthroscopy was performed in all specimens by a single experienced ankle arthroscopist with use of the same technique. The Achilles tendon and the medial and lateral malleoli were identified by palpation and marked. The level of the tibiotalar joint was estimated through palpation of the tips of the malleoli and of the joint itself while moving the foot. The approximate locations of the sural nerve and the posteromedial neurovascular bundle were estimated and drawn on the skin. Posterolateral and posteromedial portal sites immediately adjacent to the Achilles tendon at the estimated level of the tibiotalar joint were marked. The posterolateral portal was established first. Through the posterolateral portal site, an 18-gauge needle was inserted and 20 ml of normal saline solution was injected into the tibiotalar joint. The posterolateral portal was then created with use of a small longitudinal incision through the skin only. A small hemostat was used to bluntly dissect toward the lateral aspect of the joint. A 3.0-mm, 30 angled arthroscope (Dyonics; Smith and Nephew, Andover, Massachusetts) with water irrigation was then inserted into the lateral portal. Next, an 18-gauge needle was inserted through the skin at the site marked for the posteromedial portal, again oriented toward the lateral aspect of the joint. The needle was visualized through the arthroscope to ensure that it was lateral to the flexor hallucis longus tendon. The needle was removed, and the posteromedial portal was established with a small incision and blunt dissection (Fig. 1). During arthroscopy, the tibiotalar and subtalar joints, medial and lateral gutters, and flexor hallucis longus tendon were all identified. The soft tissue and capsule were debrided in all specimens to completely identify the posterior aspect of the subtalar joint and the flexor hallucis longus tendon. Electrocautery was used to mark the anterior extent of the talar dome that could be visualized with the foot manually dorsiflexed. Following the arthroscopic procedure, the metal cannulae were replaced by 4-mm plastic cannulae, which were filled with vegetable oil and were secured in the portal sites with silk suture to allow their identification on magnetic resonance imaging. A 1.5-T clinical imaging magnet (General Electric, Milwaukee, Wisconsin) with a transmit-receive quadrature extremity coil was used. T1-weighted sequences with a time to echo of 9/fraction, a time to repetition of 500 msec, and a 4.0-mm slice thickness with 0.0-mm skip per slice were obtained in the axial plane. The distances between the cannulae and the sural nerve, small saphenous nerve, posterior tibial neurovascular bundle, and flexor hallucis longus tendon were measured electronically. The results and measurements on magnetic resonance imaging were independently interpreted by two musculoskeletal radiologists. For each distance, the average of the two readings was compared with the measurement obtained during dissection. After the magnetic resonance imaging studies were performed, careful dissection was done with the tibia secured in the specimen holder. The sutures securing the cannulae were left intact over the Achilles tendon to maintain their orientation. The distances between the cannulae and the sural nerve, small saphenous vein, posterior tibial artery, tibial nerve, medial calcaneal nerve, flexor hallucis longus tendon, and Achilles tendon were measured. The measurements were made as soon as the structure was identified to minimize displacement Fig. 1 The arthroscopic instruments are placed in the right ankle joint. Note that the medial cannula remains lateral to the flexor hallucis longus tendon.

3 765 the measurements made on the magnetic resonance imaging studies was evaluated by computing the Pearson product correlation coefficient value (r). Fig. 2-A Magnetic resonance image showing the cannula adjacent to the Achilles tendon (arrow). by additional dissection. Any injury to these structures was documented. A micrometer was used to measure distances during dissection to an accuracy of 0.1 mm. The percentage of the talar dome visualized during arthroscopy was based on the distance between the posterior extent of the articular surface of the talar dome and the electrocautery mark, which had been made during arthroscopy. A paired two-tailed Student t test was used to analyze the difference between the measurements made during dissection and those made on magnetic resonance imaging studies. The interobserver correlation of Results Portal Placement ortals were created immediately adjacent to the Achilles Ptendon in all specimens. The average measured distance between the lateral cannula and the Achilles tendon was 1.0 mm (range, 0 to 5.7 mm) on dissection and 1.1 mm (range, 0 to 5.0 mm) on magnetic resonance imaging. The average measured distance between the medial cannula and the Achilles tendon was 0.6 mm (range, 0 to 5.5 mm) on dissection and 0.7 mm (range, 0 to 3.0 mm ) on magnetic resonance imaging. Of the twenty-six portals, eighteen (ten medial and eight lateral) were found to be in contact with the Achilles tendon during dissection, whereas thirteen (eight medial and five lateral) were noted to be touching the tendon on magnetic resonance imaging (Figs. 2-A and 2-B). The Achilles tendon was not damaged in any of the specimens. Posterolateral Portal Placement of the posterolateral portal was evaluated in relation to the distance from the sural nerve and the small saphenous vein. The results are summarized in Table I. The average distance between the posterolateral portal and the sural nerve was 3.2 mm (range, 0 to 8.9 mm) on dissection and 2.8 mm (range, 0 to 9.0 mm) on magnetic resonance imaging. Contact with the nerve was noted in two specimens during dissection and in three specimens on magnetic resonance imaging. The sural nerve and small saphenous vein ran adjacent to one another, with the nerve posteromedial to the vein in all but one case. Both the nerve and the vein became more anterolateral as they progressed distally through Fig. 2-B Magnetic resonance images showing the cannula displacing the flexor hallucis longus tendon medially (arrows).

4 766 TABLE I Proximity of the Posterolateral Portal to Anatomic Structures Average Distance (Range) (mm) Structure Dissection Magnetic Resonance Imaging P Value Achilles tendon 1.0 (0-5.7) 1.1 (0-5.0) 0.92 Sural nerve 3.2 (0-8.9) 2.8 (0-9.0) 0.73 Small saphenous vein 4.8 (0-11.0) 4.5 ( ) 0.77 the area where the cannula was located. The average distance between the small saphenous vein and the cannula was 4.8 mm (range, 0 to 11.0 mm) on dissection and 4.5 mm (range, 1.0 to 10.0 mm) on magnetic resonance imaging. Contact between the small saphenous vein and the cannula was identified in two specimens during dissection. There was no visible damage to the nerve or vein in any specimen. Posteromedial Portal Placement of the posteromedial portal was evaluated in relation to its distance from the flexor hallucis longus tendon and the posteromedial neurovascular bundle, including the medial calcaneal nerve. The results are summarized in Table II. The average distance between the cannula and the flexor hallucis longus tendon was 2.7 mm (range, 0 to 11.2 mm) on dissection and 2.3 mm (range, 0 to 8.5 mm) on magnetic resonance imaging. Contact between a cannula and the flexor hallucis longus tendon was identified in seven of the thirteen specimens on dissection and in five specimens on magnetic resonance imaging. Magnetic resonance imaging also demonstrated displacement of the flexor hallucis longus tendon by the cannula in four specimens (see Fig. 2-B). In all of these cases, no damage to the tendon was noted. The average distance between the posteromedial cannula and the tibial nerve was 6.4 mm (range, 0 to 16.2 mm) on dissection and 3.5 mm (range, 0 to 11.0 mm) on magnetic resonance imaging. In the cadaver specimens, the posteromedial neurovascular bundle was easily identified on magnetic resonance imaging, but the nerve and artery within the bundle could not be differentiated. Thus, the closest distance between the cannula and the identifiable neurovascular bundle was used as the distance between the cannula and the tibial nerve. The distance between the posterior tibial artery and the cannula was measured only by dissection and averaged 9.6 mm (range, 2.4 to 20.1 mm). Contact between the tibial nerve and the cannula was identified in two specimens on dissection and in two additional specimens on magnetic resonance imaging. No damage to the nerves was noted. The tibial nerve was located posterior to the flexor hallucis longus tendon in two specimens (numbers 2 and 9) on dissection and in three specimens (numbers 2, 7, and 8) on magnetic resonance imaging (Fig. 3). When the nerve is posterior to the flexor hallucis longus tendon, it is closer to the cannula than when it is located anteromedially as in the remaining specimens and as described in the literature 21. Contact between the nerve and the cannula was noted on magnetic resonance imaging in only one of the specimens (number 7) in which the nerve was located posteriorly. The medial calcaneal nerve could not be visualized on magnetic resonance imaging, and therefore measurements for it were not made. On dissection, the average distance between the cannula and the medial calcaneal nerve was 17.1 mm (range, 19 to 31 mm). Only one medial calcaneal nerve was identified in each specimen. A large difference was noted in the mean distance between the cannula and the tibial nerve (6.4 mm) and between the cannula and the medial calcaneal nerve (17.1 mm). In all but one specimen, the medial calcaneal nerve branched from the tibial nerve well distal to the TABLE II Proximity of the Posteromedial Portal to Anatomic Structures Average Distance (Range) (mm) Structure Dissection Magnetic Resonance Imaging P Value Achilles tendon 0.6 ( 0-5.5) 0.7 (0-3.0) 0.71 Flexor hallucis longus tendon 2.7 (0-11.2) 2.3 (0-8.5) 0.65 Tibial nerve 6.4 (0-16.2) 3.5 (0-11.0) 0.02 Posterior tibial artery 9.6 ( ) Medial calcaneal nerve 17.1 ( )

5 767 Fig. 3 Magnetic resonance image showing the tibial nerve (arrow) posterior to the flexor hallucis longus tendon. location of the posteromedial portal. In one specimen (number 12), however, the medial calcaneal nerve separated from the tibial nerve more proximally. Both the medial calcaneal nerve and the tibial nerve were in contact with the cannula in this specimen. No deformation or alteration of the course of the nerves was noted. Magnetic Resonance Imaging Compared with Dissection Tables I and II illustrate the consistency in measurement when magnetic resonance imaging and dissection techniques were compared. No significant difference between the measurements was found except for those of the distance between the tibial nerve and the posteromedial portal (p = 0.02). No significant interobserver difference was identified between the readings of the magnetic resonance imaging studies by the two radiologists. 1981, Drez et al. retrospectively reviewed fifty-six ankle arthroscopies performed with use of anterior and posterior portals 11. They noted that, in most cases, the posterolateral portal allows entire views of the posterior cavity and that the posteromedial portal is rarely needed. Ferkel et al. reported on what we believe is the largest consecutive series of ankle arthroscopies (612 patients) in the literature 5. The overall complication rate was 9.0%, and the rate of neurological complications was 4.4%. They recommended placement of the patient in the supine position and the use of anterolateral, anteromedial, and posterolateral portals for routine procedures. They stated that anterocentral, trans-achilles, and posteromedial portals should never be used because of their high potential for neurovascular and tendon injury. Parisien et al., in a cadaveric study, noted that posterior lesions and loose bodies within the posterior pouch were seen better and were treated more effectively through two posterior portals when the limb was in a prone position 13. Dissections during that study did not demonstrate damage to the neurovascular structures after the arthroscopic procedures. We demonstrated that an average of 54% of the talar dome can be visualized with use of the two portals described. In addition, the flexor hallucis longus tendon and sheath and the posterior aspect of the subtalar joint can be well visualized (Fig. 4). There is a general concern that the tibial nerve, the posterior tibial artery, and the medial calcaneal nerve are at risk when the posteromedial portal is used 4,15-17,19. In one of the few anatomic studies evaluating the relative safety of portal sites, Feiwell and Frey established anteromedial, anterolateral, anterocentral, posterolateral, and posteromedial portals in eighteen cadaveric specimens 16. A 3-mm arthroscope was used, and dissection was performed for each portal. They found that the Visualization of the Talar Dome The average distance between the posterior extent of the articular surface of the talar dome and the electrocautery mark, which established how far anterior the talar dome could be visualized arthroscopically, was 25.5 mm (range, 18 to 34 mm). An average of 54% (range, 42% to 73%) of the talar dome could be visualized. Discussion rthroscopy has become a valuable tool for the treatment A and diagnosis of ankle disorders. Improved techniques with use of invasive and noninvasive distraction along with new small-joint instrumentation have expanded its role 3-5,7,8,14,22. In Fig. 4 Arthroscopic view of the flexor hallucis longus tendon (arrow) and subtalar joint through the posterolateral portal.

6 768 tibial nerve was an average of 7.5 mm from the posteromedial cannula and there was one episode of contact in eighteen specimens. These findings were similar to ours, which demonstrated an average measurement of 6.4 mm and two specimens with contact between the nerve and the cannula on dissection. Feiwell and Frey reported that the average distances between the anteromedial portal and the saphenous nerve, the anterolateral portal and the superficial peroneal nerve, and the posterolateral portal and the sural nerve were 7.4 mm, 6.2 mm, and 6.0 mm, respectively 16. They concluded that the use of these portals is safe. These distances are similar to the average distance of 6.4 mm between the posteromedial portal and the tibial nerve found in our study. However, this portal still needs to be used cautiously, since an injury to this nerve would be more serious, as it could result in the loss of the vital protective sensation on the sole of the foot. The needle must remain adjacent to the Achilles tendon and must remain central during portal placement to maintain the farthest distance possible from the posteromedial bundle. This can be easily facilitated with the patient in a prone position, whereas an attempt to establish a posterior portal with the patient in the supine position results in a higher likelihood of straying from the midline. In addition, it is important to remain lateral to the flexor hallucis longus tendon during arthroscopy to protect the neurovascular bundle. In the present study, the flexor hallucis longus tendon was in direct contact with the cannula, acting as a protective barrier for the neurovascular structures entering the tarsal canal, in three of thirteen specimens. The medial calcaneal nerve was found to be an average of 2.5 mm from the posteromedial cannula in the study by Feiwell and Frey 16. They concluded that there was a substantial risk of injury to the medial calcaneal nerve with placement of the posteromedial portal. In our study, the medial calcaneal nerve branched an average of 17 mm distal to the posteromedial portal, and only one specimen had contact between the cannula and nerve. This safer distance can be explained by the positioning of the patient. With the patient supine, there is a tendency to start distal to the tibiotalar joint and to enter the joint at an angle when establishing a posterior portal 14. With the patient in the prone position, a more proximal position is maintained by visualizing the starting point directly over the tibiotalar joint. In the literature, the posterolateral portal is considered safe 13-17,19,22. We found that the average separation between the sural nerve and the posterolateral portal was 3.2 mm, which is closer than the 6.0-mm separation found by Feiwell and Frey 16. This discrepancy can be explained by the difference between the starting position when the limb was in the supine position and when it was in the prone position. We found the sural nerve to lie more anterolaterally as it progressed distally past the cannula. This resulted in a greater separation between a more distally placed portal and the nerve when the limb was in the supine position than would occur with a more proximally placed portal with the limb in the prone position. The present investigation was a cadaveric study with its associated weaknesses and flaws. Measurements obtained during dissection could be influenced by movement of the cannulae and the anatomic structures. Although gross nerve injury from the cannula can be identified on dissection, an anatomic study alone cannot detect all nerve injuries. This obviously is a weakness of the study in terms of determining clinical efficacy. Magnetic resonance imaging was used as an additional measurement to confirm the dissection results. No significant differences in the measurements were detected, with the exception of the distance between the posteromedial cannula and the tibial nerve. Several factors contributed to this difference. As previously noted, the radiologists could not identify the individual structures within the posteromedial neurovascular bundle. Therefore, the closest distance from the cannula to the neurovascular bundle was used as a surrogate for the distance to the tibial nerve. Furthermore, magnetic resonance images were not perpendicular to the neurovascular structures and the cannulae in all specimens. This may have altered the measured distances. Also, the positioning of the cadaver foot and ankle within the extremity coil of the magnetic resonance imaging unit may have caused movement of the plastic cannulae. We did not test the reproducibility of the magnetic resonance imaging measurements after removing and reinserting the cannulae. Regardless, the measurements on magnetic resonance imaging were consistently more conservative than the measurements on dissection, and they demonstrated a high interobserver correlation. Because dissection was not required, distortion and movement of the anatomic structures were avoided and we believe that magnetic resonance imaging was a valuable tool in the evaluation of the anatomic relationships during this study. The present cadaveric study suggests that, during posterior ankle arthroscopy with the limb in the prone position, the posteromedial and posterolateral portals could be used with a relatively small risk to the neurovascular structures, if the technique for portal placement is followed cautiously. This approach affords visualization and accessibility to the posterior half of the tibiotalar joint, subtalar joints, and the flexor hallucis longus tendon and its sheath. On the basis of this anatomic study, it seems reasonable to proceed with limited clinical trials to verify that critical neurovascular structures would not be injured if arthroscopy through posterior portals were performed on patients in the prone position. David F. Sitler, MD Navy Medical Center San Diego, San Diego, CA address: dsitler@nmcsd.med.navy.mil Annunziato Amendola, MD Department of Orthopaedic Surgery, University of Iowa Hospitals and Clinics, Iowa City, IA address: ned-amendola@uiowa.edu Christopher S. Bailey, MD Fowler Kennedy Sport Medicine Clinic, 3M Centre, University of Western Ontario, London, ON N6A 3K7, Canada

7 769 Lisa M.F. Thain, MD Alison Spouge, MD Department of Radiology, University of Western Ontario, London, ON N6A 3K7, Canada In support of their research or preparation of this manuscript, one or more of the authors received a Canadian Orthopaedic Foundation Research Grant. None of the authors received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the authors are affiliated or associated. References 1. Burman MS. Arthroscopy or the direct visualization of joints. An experimental cadaver study. J Bone Joint Surg. 1931;13: Martin DF, Baker CL, Curl WW, Andrews JR, Robie DB, Haas AF. Operative ankle arthroscopy. Long-term followup. Am J Sports Med. 1989;17: Amendola A, Petrik J, Webster-Bogaert S. Ankle arthroscopy: outcome in 79 consecutive patients. Arthroscopy. 1996;12: Ferkel RD, Fasulo GJ. Arthroscopic treatment of ankle injuries. Orthop Clin North Am. 1994;25: Ferkel RD, Heath DD, Guhl JF. Neurological complications of ankle arthroscopy. Arthroscopy. 1996;12: Barber FA, Click J, Britt BT. Complication of ankle arthroscopy. Foot Ankle. 1990;10: Guhl JF. New concepts (distraction) in ankle arthroscopy. Arthroscopy. 1988;4: Guhl JF, Stone JW, Whit Ewing J. Ankle arthroscopy: special equipment, operating room setup, and technique. In: McGinty JB, Caspari RB, Jackson RW, Poehling GG, editors. Operative arthroscopy. 2nd ed. New York: Raven Press; p Buckingham RA, Winson IG, Kelly AJ. An anatomical study of a new portal for ankle arthroscopy. J Bone Joint Surg Br. 1997;79: Saito A, Kikuchi S. Anatomic relations between ankle arthroscopic portal sites and the superficial peroneal and saphenous nerves. Foot Ankle Int. 1998;19: Drez D Jr, Guhl JF, Gollehon DL. Ankle arthroscopy: technique and indications. Foot Ankle. 1981;2: Andrews JR, Previte WJ, Carson WG. Arthroscopy of the ankle: technique and normal anatomy. Foot Ankle. 1985;6: Parisien JS, Vangsness T, Feldman R. Diagnostic and operative arthroscopy of the ankle. An experimental approach. Clin Orthop. 1987;224: Guhl JF. Foot and ankle arthroscopy. 2nd ed. Thorofare, NJ: Slack; Portals; p Voto SJ, Ewing JW, Fleissner PR Jr, Alfonso M, Kufel M. Ankle arthroscopy: neurovascular and arthroscopic anatomy of standard and trans-achilles tendon portal placement. Arthroscopy. 1989;5: Feiwell LA, Frey C. Anatomic study of arthroscopic portal sites of the ankle. Foot Ankle. 1993;14: Ferkel RD. Complications in ankle and foot arthroscopy. In: Ferkel RD, Whipple TL, editors. Arthroscopic surgery: the foot and ankle. Philadelphia: Lippincott-Raven; p Guhl JF, Patel D. Arthroscopic anatomy. In: Guhl JF, editor. Foot and ankle arthroscopy. 2nd ed. Thorofare, NJ: Slack; p Morgan CD. Gross and arthroscopic anatomy of the ankle. In: McGinty JB, Caspari RB, Jackson RW, Poehling GG, editors. Operative arthroscopy. 2nd ed. New York: Raven Press; p Zimmer T, Ferkel RD. Future developments. B. Endoscopic procedures for the retrocalcaneal bursa, plantar fascia, and achilles tendon. In: Ferkel RD, Whipple TL, editors. Arthroscopic surgery: the foot and ankle. Philadelphia: Lippincott-Raven; p Sarrafian SK. Anatomy of the foot and ankle: descriptive, topographic, functional. 2nd ed. Philadelphia: Lippincott; Ferkel RD. Diagnostic arthroscopic examination. In: Ferkel RD, Whipple TL, editors. Arthroscopic surgery: the foot and ankle. Philadelphia: Lippincott- Raven; p