A dissection and computer tomograph study of tarsal coalitions in 100 cadaver feet

Transcription

1 ELSEVIER Journal of Orthopaedic Research 21 (2003) Journal of Orthopaedic Research hres A dissection and computer tomograph study of tarsal coalitions in 100 cadaver feet L.B. Solomon F.J. Riihli b,c, J. Taylor ', L. Ferris d, R. Pope d, M. Henneberg a Departn?ent of' Orthopuedics. Alice Springs Ho.~pitul, F1iniier.s University School of' Meu'icirie, P. 0. Box 2234, Alice Springs. NT 0871, Australia Depurtnient of' Anatomicul Scierzwr.. Uniwrsity of Adelaide, Adidaide, SA 5005, Austruliu Depurrrnent of Rudioliig.~, Royd Adeluicle Hospital. Adduiilc~. SA 5000, Amtraliu Wakefield Orthopuerlic Clinic, Adeluirie, SA 5000, Au.str.uliu Received 28 September 2001: accepted 25 July 2002 Abstract Most of the clinical studies report the incidence of tarsal coalitions () as less than 1% but they disregard the asymptomatic coalitions. have been associated with degenerative arthritic changes. After X-rays, computer tomography (CT) is the most commonly used diagnostic test in the detection of. The aims of our study were to establish the incidence of ; the association between and accessory tarsal bones and between and tarsal arthritis; and to assess the sensitivity of CT as a diagnostic tool in. We performed spiral CT scans of 100 cadaver feet (mean age at death 77.7 & 10.4), which were subsequently dissected. The dissections identified nine non-osseous : two talocalcaneal and seven calcaneonavicular. There was no osseous coalition. Tarsal arthritis was identified in 3 1 cases. Both talocalcaneal coalitions were associated with arthritis while none of the calcaneonavicular coalitions were associated with tarsal arthritis. The CT diagnosed an osseous talocalcaneal coalition and was suspicious of fibrocartilaginous coalitions in eight cases. There was correlation between dissection and CT in two talocalcaneal coalitions and three calcaneonavicular coalitions thus CT identifying 55.50/0 of the coalitions. CT did not diagnose four non-osseous coalitions and diagnosed errouresly four possible coalitions. In conclusion our study demonstrated that the incidence of non-osseous is higher than previously thought ( 12.72'%). The calcaneonavicular coalitions are the most common single type (9.09%) and they do not seem to be associated with arthritic changes in the tarsal bones. Our CT results suggest that spiral CT has a low sensitivity in the detection of non-osseous coalitions and questions if multislice CT should be used routinely when are suspected. Crown Copyright Published by Elsevier Science Ltd. on behalf of Orthopaedic Research Society. All rights reserved. Introduction Tarsal coalitions () are defined as fibrous, cartilaginous or osseous union of two or more tarsal bones [6,10]. Due to the irregular shape of tarsal joints and the superposition of neighboring bones it can be difficult to demonstrate coalitions on standard X-rays. Computer tomography (CT) is considered the preferred diagnostic imaging technique for talocalcaneal coalitions [l I, 141. Even if most of the calcaneonavicular coalitions can be identified on oblique X-rays, CT scans can bring further *Corresponding author. Tel.: ; fax: , arltlrrss: bogdan2500@yahoo.com (L.B. Solomon). details and confirm other conditions and additional changes [3]. The incidence of was most often reported as less than 1% [3,14]. All the clinical studies will miss at least the asymptomatic coalitions and Leonard [8] found that 76% of subjects with are symptom free. Most of the are diagnosed as they become symptomatic and most coalitions become symptomatic when they ossify and the patient looses motion in the subtalar joint complex. This is also the cause for abnormal stresses, which will eventually lead to secondary degenerative changes [14]. By performing a comparative dissection and CT study in cadavers we aimed to establish the incidence of, the association between and accessory tarsal bones, the association between and tarsal arthritis and the sensitivity of CT as a diagnostic tool in /03/$ - see front matter Crown Copyright Published by Elsevier Science Ltd. on behalf of Orthopaedic Research Society. All rights reserved. PI1: S (02)

2 L. B. Soloriton et ul. I Jourml of Oithopui& Re.warch 21 (2003) Methods We used 100 feet from 55 consecutive embalmed cadavers for the Ray Last Laboratory (dissecting room) of the Department of Anatomical Sciences of the University of Adelaide. There were 45 paired feet and 10 unpaired feet (the other 10 feet were excluded because their anatomy was altered by student dissection prior to the beginning of our study). No past medical history of the subjects was known but all feet had a normal appearance and there were no scars to suggest previous surgery. The mean age at death was 77.7 (sd. 10.4) years with the range from 43 to 101 years. After labeling, the feet were placed in a support and subjected to fine (1 mm) coronal and axial CT scans. Helical scanning was done in the Department of Radiology of the Royal Adelaide Hospital on a Picker PQ 6000 scanner. The post-processing was done on a Voxel Q workstation and reconstruction in arbitrary planes (curved multiplanar reconstruction) was performed for questionable cases. All scans were read by the two radiologists involved in the study (JT and FJR). The overall bone density was assessed on a subjective scale of 1 to 5 with 1 signifying a low density and 5 an abnormal high density. Any coalition was recorded as well as its location. size and structure. Osseous coalitions were diagnosed by the presence of an osseous bar joining adjacent tarsal bones. Fibrocartilaginous coalitions were considered as suspicions when there was marked narrowing of the space between the two bones associated with interruption of the subchondral bone and subchondral sclerosis and cysts. We also recorded the subtalar joint space width and the space between the calcaneus and the navicular on a scale from 1 to 5, where 1 (1 mm) is narrow, 3 (3 mm) normal and 5 (5 mm) wide. The reactive osteoarthritic changes (subchondral sclerosis, osteophytes and rarefaction) were noted in type, location and severity. We also noted the number and location of accessory ossicles. After scanning, the feet were dissected out and once again the, the accessory tarsal bones and arthritic changes were noted. All the soft tissue was removed piecemeal until only the bony structures and their fibrous andlor tendinous interconnections were left in place. At the end all the joints were opened and the pattern of articular surfaces was recorded as well as the disposition of the intertarsal synovial joints. For the we recorded their location, constitution (osseous or non-osseous) and dimensions. Every bony structure identified by the presence of trabecular bone, besides the seven tarsal bones was recorded as an accessory tarsal bone. Because the use of cadaveric feet and the impossibility of diagnosis early cartilaginous arthritic changes in non-fresh and embalmed specimens, only grade 111 (loss of partial thickness) and IV (exposure of subchondral bone) chondral defects [9] were documented as arthritis in the subtalar joint complex. Because of the impossibility of uniform assessment of other signs of arthritis (like osteophytes) no other arthritic changes were documented on dissection. After the dissection and CT findings were compared the two radiologists for all cases that had a dissection and/or C T diagnosis of coalition on their first reading performed two more CT readings. Results No signs of trauma, infection, neoplasm or other causes of secondary coalitions could be found on the CTs or dissection of the feet. The dissection of the 100 feet identified nine non-osseous fibrocartilaginous (Figs. 1 and 2). Seven coalitions were calcaneonavicular and two were talocalcaneal (Table 1). The seven-calcaneonavicular coalitions were identified in feet from five subjects, in two cases the coalitions being bilateral. Their size varied from 5 mm x 15 mm to 15 mm x 35 mm. The large 15 mm x 35 mm coalitions were identified in both feet of the same subject (Fig. 3). The two talocalcaneal coalitions were identified in feet from two subjects. One of the coalitions was involving the entire middle talocalcaneal articular surface. The other coalition was extraarticular, located anterior to the posterior talocalcaneal joint but posterior to sinus tarsi and lateral to the anterior and middle talocalcaneal articular surfaces and their capsule (Fig. 4). Both coalitions had a cross section of approximately 1 cm2. We identified eight ossa trigonum, 31 sesamoid bones in the tendon of tibialis posterior and 17 sesamoid bones in the tendon of fibularis longus. Tarsal arthritis was identified in 31 cases. Both talocalcaneal coalitions were associated with talocalcaneal arthritis while none of the calcaneonavicular coalitions was associated with dissection findings of tarsal arthritis. CT diagnosed or was suspicious of a coalition in nine feet from eight subjects. As presented in Table 1 in five of these cases the CTs were considered positive for coalitions in all readings, while in the other four cases the Fig. I. Specimen 71. Talocalcaneal coalition (arrow). N-navicular; C--cakaneus

3 354 L.B. Solomon et ul. I Journal of Orthopaedic Research 21 (2003) Fig. 2. Specimen 71. Micro CT. Arrows-site of talocalcaneal coalition. There are no osseous trabeculae crossing the coalition. The coalition space is smaller than the intertrabecullar space. N-navicular; C-calcaneus. CTs were considered positive in one reading and negative in the other readings. There was one osseous talocalcaneal coalition, two suspicions of fibrocartilaginous talocalcaneal coalitions, four suspicions of fibrocartilaginous calcaneonavicular coalitions, one suspicion of a fibrocartilaginous calcaneocuboid coalition and one suspicion of a fibrocartilaginous cuboidocuneiform coalition. There was dissection CT correlation in two talocalcaneal and in three calcaneonavicular coalitions (Figs. 1 and 5). One of the calcaneonavicular coalition shown by dissection had a CT diagnosis of suspicion of calcaneocuboid coalition. There was no CT diagnosis of coalition in the other three calcaneonavicular coalitions found during dissections. The dissections did not confirm the CT diagnosis of an osseous talocalcaneal coalition and of a suspicion of a fibrocartilaginous calcaneonavicular coalition and a cuboidocuneiform coalition. Beside the three cases, included in the above nine, where the CT was positive for coalition in all readings but there was no dissection confirmation of coalition we had other five cases in which the initial CT reading suspected a coalition that was not confirmed at subsequent readings nor at dissections. The subtalar joint width was narrowed in 27%. The width of the joint was measured as 4 mm in 1/100, 3 mm in 62/100 and 2 mm in 27/100. The calcaneonavicular space was measured 1 mm in 25/100, 2 mm in 33/100, 3 mm in 25/100, 4 mm in 10/100 and 5 mm in 5/100 with an average of 2.3 mm. Osteoarthritic changes were described in 36/100 cases. Three of the cases with osteoarthritic changes diagnosed on CT in the subtalar joints were in specimens with dissection f CT diagnosis of calcaneonavicular coalitions (none of these had confirmed arthritic changes on dissection). CT also identified 24 ossa trigonum, 16 sesamoid bones in the tendon of fibularis longus and three sesamoid bones in the tendon of tibialis posterior. Discussion Discovered last century by anatomists (cited by Pachuda [12]), came to clinicians attention when Table 1 The dissection and CT findings of in the 100 feet investigated Nr. Specimen Coalition on Coalition on CT- Coalition on CT-- Coalition on CT-- Dissection diagnosis CT diagnosis dissection first reading second reading third reading of arthritis of arthritis Specimens positiae for at dissection l-t CT C V Cub-cun Calc-cub Cub-cun Specimens positive for at Jirst CT reading only talocdicaneal coalition; 4akdneonavicular coalition; calc-cub-calcaneocuboid coalition; cub-cun-cuboideocuneiform coalition. The italicised rows mark the specimens with coalitions confirmed on both CTs and dissection. Calc-cub - -

4 L. B. Solonion et al. I Journal of Ortliopaedic Riwarch 21 (2003) Fig. 3. Specimens 67 and 69. Bilateral calcaneonavicular coalition (arrows). The coalition was opened in specimen 67 Fig. 4. Specimen 58. Lateral view of the talocalcaneal complex. Extraarticular talocalcaneal coalition (arrow heads). The arrows point the posterior talocalcaneal joint. they were associated with the flat foot and the peroneal spasm at the beginning of last century [1,15]. However, their clinical expression is variable and the issue of muscle spasm in is still controversial. The incidence of is not known. Most of the studies report it as less than 1!40 but they disregard the asymptomatic coalitions. If we consider Leonard's study [8] (who investigated 98 first-degree relatives of 31 patients with symptomatic ) that concluded that 76% of the coalitions are asymptomatic the incidence could be 4%. Harris and Beath [5] examined 3600 recruits and found the incidence of peroneal spastic flat foot to be 2%. They mention only one case of calcaneonavicular bar demonstrated radiologically but make no comment about how many of the cases had been X-rayed. Vaughan and Segal [19] performed approximately 2000 X-ray examinations for painful feet in an Army Hospital and demonstrated 21 (~1%). Stormont and Peterson [ 161 described that of 32 cases with calcaneonavicular coalitions, which is thought to be best detected on oblique X-rays and not on CT [6], 10 could not be identified radiologically but only at the time of surgery.

5 356 L. B. Solomon et al. i Journtrl oj 0rtIiopuc.clic Reseutdi 21 (2003) Fig. 5. Specimen 7 I. Calcaneonavicular coalition (arrow) confirmed by dissection According to the study, the incidence of calcaneonavicular coalitions undiagnosed radiologically could be 30%. The most accurate reported incidence of the should be one by Pfitzer in Unfortunately, the results published in Pfitzer s paper are cited differently by different authors: Mosier and Asher [lo] stated that Pfitzer s study was performed on 520 foot skeletons and the incidence of calcaneonavicular coalitions was 2.9% and the global one 6%; Leonard [8] stated that the study involved 524 feet and the incidence of identified was 0.38% (2 feet); Badgley [l] stated that Pfitzer s study involved 840 feet and the incidence of calcaneus secundarius (calcaneonavicular coalition) was I %l (nine cases). Harris [4] examined 20 embryos (40 feet) and reported six talocalcaneal coalitions in four embryos (20%) and no calcaneonavicular coalitions. She also reported tarsometatarsal coalition and accessory ossicles totalling 50% anatomical variations in apparently normal embryos and questions how many of these persist in the definitive foot. Our dissections demonstrated that the incidence of is , higher than in all previous reports (excepting Harris s [4], but of course she did not report the incidence but the mechanism). The incidence of calcaneonavicular coalitions was 9.09% (seven coalition in five of 55 subjects), and the incidence of talocalcaneal coalitions was 3.63% (two coalitions in two of 55 subjects). The CT results of our study suggested an incidence of 14.54% (nine coalitions in eight of 55 subjects). There was dissection-ct correlation in five cases (two talocalcaneal and three calcaneonavicular coalitions). There was no CT diagnosis of coalition in three of the calcaneonavicular coalitions found during dissection, including the two large 15 mm/35 mm coalitions. In one case of calcaneonavicular coalition the CT was suspicious of a cuboideonavicular coalition. The other foot of this last specimen had a calcaneonavicular coalition identified on both CT and dissection. In three cases the dissections did not confirm the CT diagnosis of coalition (one osseous talocalcaneal, one calcaneonavicular and one cuboideocuneiform). Although difficult to quantify, these results would suggest that the sensitivity of spiral CT is not very high in the detection of non-osseous coalitions. We believe that this is mainly related to the difficulty in establishing a CT diagnosis for non-osseous coalitions given the fact that the CT criteria for are very similar to the CT criteria for osteoarthritis [ll] (Fig. 6). From this point of view we can comment that in one case of calcaneonavicular coalition with no CT confirmation, the CT suspected a calcaneocuboid coalition and in other two cases of calcaneonavicular coalitions there was a CT diagnosis of degenerative changes. This is not the first report regarding the low sensitivity and limitations of standard spiral CT in diagnosing non-osseous and we question if multislice CT should not be used routinely instead? Wechsler et al. [20] compared the CT and MRI in the detection and characterisation of nine. In that study, CT depicted six coalitions out of nine and only four of the six were characterised correctly. None of the fibrous coalitions were characterised correctly. MRI depicted all coalitions and seven were characterised correctly. MRI was also false positive in a case of synovitis. The main significance of the higher incidence of is that probably more than expected are asymptomatic. An incidence of /0 makes the non-osseous a rather common anatomical variant.

6 L. B. Solomon rt (11. I Journal of Ortliopprreclic Research 21 (2003) Fig. 6. Specimen 75. CT diagnosis of arthritis. Arrows--subcortical cysts. Dissection demonstrated a calcaneonavicular coalition. The paired foot of this specimen, specimen 71 had a both CT and dissection diagnosis of calcaneonavicular coalition. Limitation of movement was reported to be the cause for the symptomatic foot in and for the secondary arthritis [14]. In accordance with this it would appear that small non-osseous coalitions would allow enough movement in the subtalar joints to prevent the coalition from becoming symptomatic and also to prevent secondary arthritic changes. The same would apply to explain why the coalitions do not ossify. The issue is still controversial and for example Kumai et al. [7] present an interesting explanation after a histopathological study: incomplete coalitions produce microfractures and remodeling at the boundary between bone and coalition, which then lead to degenerative changes. These in turn induce pain via the free nerve endings in the periosteum and articular capsule surrounding the coalition. One might speculate that the same mechanism could apply for small osseous coalitions. It is known that the large multiple coalitions associated with the ball and socket ankle are symptom free and do not produce secondary degenerative changes [ 181. None of the calcaneonavicular coalitions identified on our dissections had cartilage change of arthritis but in three of these specimens there were CT changes of arthritis. A statistical issue needs to be raised about the incidence of. We have identified in feet from seven out of 55 subjects, which brings the incidence of to 12.72% f 4.5. In the mean time we have identified in nine out of 100 feet. From a statistical point of view the first incidence (of 12.72%) is correct because a person is counted as having irrespective of the fact that he has one or two feet involved. From a practical point of view, as only symptomatic coalitions require treatment, it might be more appropriate to refer to the incidence of coalitions reported in the number of single feet. In this second case, as not all coalitions are bilateral, the incidence is smaller (9 %1), but not statistically significantly different from that found per person. Our numbers are to small to make a pertinent comment about the percentage of bilateral and different types of coalition but the 40 % incidence of bilateral calcaneonavicular coalitions is similar to other reports [6,14]. We found calcaneonavicular coalition more frequent (71.4 X)) then other recent reports that say that the proportion between the two main types of coalitions is about the same [6,14]. Excepting the secondary coalitions (post-infections, etc.) two theories have been advanced regarding their etiology. The first belongs to Pfitzer (cited by Badgley [l], Stormont and Peterson [16], Mosier and Asher [lo]) who, secondary to his dissection, hypothesized that result by incorporation of accessory bones into the nearby tarsals. Pfitzer s theory was considered failed when Harris [4] demonstrated in foetus. She produced the second theory, stating that coalitions occur as a result of failure of differentiation and segmentation of foetal mesenchyme. Without rejecting Harris s theory one can note that her work involved 20 embryos and that the embryos described with talocalcaneal coalitions were aged 25, 27.8, 60.9 and 72.3 mm. This paper was only published as an abstract and there is no mention how the embryos length was measured. The reported length of the embryos suggest that the embryos were aged between 9 and 13 weeks (estimation according to Rumack et al. [13]). It is known that the joints of the

7 358 L. B. Solomon er ul. I Journul of Ortliopuedic Research 21 (2003) foot develop after the 7th week. As demonstrated by Drachman [2], joint space cavitation occurs only after active movements take place. As a consequence, we question if the differentiation of the mesenchyme can take place in the talocalcaneal area at a later age. The other question, that even Harris [4] raised, is how many of these variants would persist in the definitive foot and how many represent a temporary developmental phase. The cause of the coalition remains unknown, but strong hereditary components have been demonstrated by several authors [8,22]. As most of the accessory ossicles that we have identified were actually sesamoid bones located at places were the tibialis posterior and the peroneus longus hook around the navicular or the cuboid and change direction, and are therefore directly related to the function of these tendons, our study could not make any correlation between the and the accessory tarsal bones. In appreciating the significance of this condition it is interesting to note that there are literature reports of symptomatic that become asymptomatic in time [17]. Also, even if the therapy options have changed in time, as more data and experience were accumulated, good results have been reported with each technique despite the fact that some of them have different philosophies (e.g. fusion versus resection). Most of the studies mention the presence of arthritic changes as the most important factor in deciding if the resection of a coalition is beneficial. The study of Wilde et al. [21] is the only report to compare the results of resection (talocalcaneal) in the presence or absence of objectively determined signs of arthritis as assessed on CT scans. This study confirmed that resection of coalitions gives good results when coalitions are not associated with arthritis. It is interesting to note that the presence of signs of arthritis correlated with the extension of the coalition to more than 50% of the joint space (Wilde et al. [21]). Acknowledgements The CT scanning was funded from the research fund of the Department of Radiology of Royal Adelaide Hospital. References [l] Badgley CE. Coalition of the calcaneus and the navicular. Arch Surg 1927;15: [2] Drachman DB. Normal development and congenital malformation of joints. Bull Rheum Dis 1969;19: [3] Drennan JC. Tarsal coalitions. Instr Course Lect 1996;45: [4] Harris BJ. Anomalous structures in the developing human foot. Anat Rec 1955;121:399. [5] Harris RI, Beath T. Etiology of peroneal spastic flat foot. J Bone Joint Surg [Br] 1948;30: [6] Kulik SA, Clanton TO. Tarsal coalition. Foot Ankle Int 1996; 17: [7] Kumai T, Takakura Y, Akiyama K, Higashiyama I, Tamai S. Histopathological study of nonosseous tarsal coalition. Foot Ankle Int 1998; I9:525-3 I. [8] Leonard MA. The inheritance of tarsal coalition and its relationship to spastic flat foot. J Bone Joint Surg [Br] 1974;56: [9] Miller I11 RH. Knee injuries. In: Canale ST, editor. Campbell s operative orthopaedics. 9th ed. St. Louis: Mosby; p [lo] Mosier KM, Asher M. Tarsal coalition and peroneal spastic flat foot. A review. J Bone Joint Surg [Am] 1984;66(111): [I I] Newman JS, Newberg AH. Congenital tarsal coalition: multimodality evaluation with emphasis on CT ad MR imaging. Radiographics 2000;20: [12] Pachuda NM, Lasday SD, Jay RM. Tarsal coalition: etiology, diagnosis. and treatment. J Foot Surg 1990;29: [I31 Rumack CM, Wilson SR, Charboneau JW. In: Diagnostic ultrasound. 2nd ed. St. Louis: Mosby; p [I41 Sakellariou A, Claridge RJ. Tarsal coalition. Orthopedics 1999; 22: [15] Slomann HC. On coalitio calcaneo-navicularis. J Orthop Surg 1921 ;3: [16] Storniont DM, Peterson HA. The relative incidence of tarsal coalition. Clin Orthop 1983;181: [I71 Sullivan JA. Tarsal coalition. In: Morrisy RT, editor. Lovell and Winter s pediatric orthopaedics. Philadelphia: Lippincott; p [IS] Takakura Y, Tanaka Y, Kumai T, Sugimoto K. Development of the ball-and-socked ankle as assessed by radiography and arthrogrdphy. J Bone Joint Surg [Br] 1999;81: [19] Vaughan WH, Segdl G. Tarsal coalition, with special reference to roentgenographic interpretation. Radiology 1953;60: [20] Wechsler RJ, Schweitzer ME, Deely DM, Horn BD, Pizzutillo PD. Tarsal coalitions: depiction and characterization with CT and MR imaging. Radiology 1994;193: [21] Wilde PH. Torode IP, Dickens DR, Cole WG. Resection for symptomatic talocalcaneal coalition. J Bone Joint Surg [Br] 1994; 76: [22] Wray JB, Herndon. Hereditary transmission of congenital coalition of the calcaneus to the navicular. J Bone Joint Surg [Am] 1963;45: