Peritalar dislocations: a retrospective study of 18 cases

Peritalar Dislocations: A RetrospectiveStudy of 18 Cases

Raffaele Garofalo, MD,1 Biagio Moretti, MD,2 Vito Ortolano, MD,3 Pasquale Cariola, MD,4

Giuseppe Solarino, MD,5 Michael Wettstein, MD,6 and Elyazid Mouhsine, MD7

The purpose of this study is to retrospectively evaluate 18 consecutive cases of peritalar dislocationsreferred to our department during a period of 25 years and to delineate the factors influencing long-termprognosis. There were 13 (73%) medial and 5 (27%) lateral dislocations. Six patients (33%) suffered anopen injury, including 2 of 13 (15%) medial and 4 of 5 (80%) lateral dislocations. Associated fracturesinvolving the hindfoot or forefoot were noted in 7 feet, including 3 of 5 lateral dislocation cases.Reduction was accomplished under general anesthesia; in no case was open reduction necessary. In 4of 6 open injuries with associated fractures, temporary fixation with Kirschner wires was performed.Patients were immobilized in a plaster cast for 4 weeks, or for 6 weeks in the presence of fracture,followed by weightbearing as tolerated. At a mean follow-up of 10.2 years (range, 4 to 26 years), 10patients (56%) showed excellent results; all had sustained a closed medial low-energy dislocation. Therewere 3 cases (17%) with fair results and 5 cases (28%) with poor results. Forty-five percent of patientsshowed a restriction of activity, a reduction of subtalar range of motion, and moderate or severeradiographic signs of hindfoot degenerative arthritis. There were no cases of talar avascular necrosis, andin no case was secondary surgery necessary. Lateral dislocation and open medial dislocations withconcomitant fractures showed a greater potential for poor prognosis. The results were independent fromperiod of cast immobilization, suggesting that 4 to 6 weeks of immobilization provides acceptablelong-term results. (The Journal of Foot & Ankle Surgery 43(3):166-172, 2004)

Key words: dislocation, subtalar, peritalar, hindfoot, degenerative arthritis

juryalareckion

ofeenbtalten

ight,inedtudy

tionMe-for

talarpos-

iants isona-ca-

teralriesan

sus-k ofointacesaluss andfootna-rior

nt ofgnon.chand

Bari,

Bari,

dics,

Bari,

ver-

logy

ons

Peritalar dislocation is a term used to describe an ininvolving a simultaneous dislocation to both the subtand the talonavicular joints without a fracture of talar nor disruption of the tibiotalar relationship. This dislocatis uncommon, accounting for approximately 1% to 1.5%all traumatic foot injuries (1). These injuries have also breferred to as subtalar or subastragalar dislocation, supedis luxatio, or hindfoot dislocation (2–5). They of

Address correspondence to: Elyazid Mouhsine, MD, DepartmeOrthopaedic and Traumatology, University of Lausanne, Rue du Bu46, 1011 Lausanne, Switzerland. E-mail: elyazid.mouhsine@hospvd

1Orthopedic Surgeon, Chief of Clinic, Department of OrthopedicTraumatology, University of Lausanne, Lausanne, Switzerland.

2Associate Professor, Department of Orthopedics, University ofBari, Italy.

3Senior Resident, Department of Orthopedics, University of Bari,Italy.

4Orthopedic Surgeon, Chief of Clinic, Department of OrthopeUniversity of Bari, Bari, Italy.

5University Researcher, Department of Orthopedics, University ofBari, Italy.

6Chief of Clinic, Department of Orthopedic and Traumatology, Unisity of Lausanne, Lausanne, Switzerland.

7Associate Professor, Department of Orthopedic and TraumatoUniversity of Lausanne, Lausanne, Switzerland.

Copyright © 2004 by theAmerican College of Foot and Ankle Surge1067-2516/04/4303-0006$30.00/0
doi:10.1053/j.jfas.2004.03.008
166 THE JOURNAL OF FOOT & ANKLE SURGERY

o

result from high-energy trauma, such as a fall from a hebut can be the result of sports injuries. Grantham (6) cothe term basketball foot, because 4 of 5 patients in his swere injured while playing basketball.

The terminology used to describe peritalar dislocadescribes the direction of peritalar foot displacement.dial dislocation is the most frequent type, accountingapproximately 80% of reported cases (5). Lateral peridislocations are less common (17%), and anterior andterior dislocations are infrequent (3%) (7, 8). A rare varinvolves total dislocation of the talus, in which the talucompletely dislocated from the ankle, subtalar, and talvicular joints (5). Given that anterior and posterior dislotions always exhibit some degree of medial and ladisplacement, it is suggested that only 2 groups of injuexist: medial and lateral (9). In medial dislocations,inversion force is applied to a plantarflexed foot. Thetentaculum tali act as a fulcrum around which the nectalus pivots, causing dislocation of the talonavicular jfirst followed by the subtalar joint. The calcaneus displmedial to the long axis of the leg, and the head of tbecomes lodged between the extensor hallucis longuthe long-toe extensors. The clinical position of theresembles a clubfoot deformity. In lateral injuries, a protion force is applied to a plantarflexed foot, with the ante

,

calcaneal process acting as a fulcrum for rotation of the

anterolateral corner of the talus. The head of talus is forcedthrough the talonavicular capsule, and becomes palpableover the medial aspect of the foot, while the calcaneus anddistal part of the foot dislocate laterally. The clinical ap-pearance is that of a flatfoot (10).

The final outcome appears to depend on the magnitude ofthe original impact (5). Although in vitro studies haveshown the extent of ligamentous damage to be similar inboth lateral and medial dislocations, lateral dislocationshave been found to carry a less favorable prognosis becausethey frequently result from higher injury trauma and arecommonly associated with fractures and open wounds (9–12). The aim of present study is to report the functional andradiographic outcome after peritalar dislocations and toattempt to delineate the factors that contribute to prognosis.

Materials and Methods

From 1973 to 1998, 21 consecutive adult patients withperitalar dislocation of the foot were referred to and treatedat our institution. All patients were transported to our centerdirectly from the scene of the accident. At follow-up, 2patients were lost, 1 patient had died, and the remaining 18were available for evaluation. These 18 patients were inter-viewed and clinically examined by the third author, whowas blinded to the type of dislocation suffered by patient.The evaluation of final outcome included an assessment ofthe following criteria: the patient’s level of daily and rec-reational activity, the degree of pain (no pain, occasionalpain, or pain during all activities), the presence of a limp,the necessity for a brace or special foot orthoses, subtalarinstability, and the range of subtalar joint motion. Range ofsubtalar joint motion was evaluated by using the methoddescribed by Inman and Mann (13) and compared with thecontralateral foot. Anteroposterior and lateral radiographswere also performed and evaluated by the sixth author(M.W.), who was blinded about type of injury sustained bythe patient. Emphasis was placed on the signs of peritalardegenerative arthritis, a decrease in joint space, subchondralsclerosis, or osteophyte formation in the subtalar or talona-vicular joints.

We considered an excellent result to be those patientswith no pain or limp, unrestricted daily and recreationalactivities, nearly complete subtalar motion, an absence ofsubjective instability (recurrent giving way of the hindfoot),and absent or minimal signs of radiographic abnormalities.A fair result was defined as those patients with occasionalpain or discomfort, especially when walking on an unevensurface; a partial restriction of daily and recreational activ-ities; up to a 50% restriction of subtalar motion; subjectiveand objective instability of the hindfoot; and the presence ofmoderate radiographic signs of degenerative arthritis of the
subtalar and/or talonavicular joints. Patients who com-
plained of pain during all activities, showed a severe limi-tation of daily and recreational activities, required the use ofan ankle brace or special foot orthosis, lost more than 50%of subtalar motion, and exhibited hindfoot instability andsevere radiographic signs of degenerative changes to thesubtalar and/or talonavicular joints were classified as poorresults.

Results

Table 1 provides a summary of patient data. Patients werefollowed up at an average of 10.2 years (range, 4 to 26years). The average age of patients at the time of injury was34 years (range, 18 to 59 years). There were 11 men and 7women; 10 dislocations involved the right foot and 8 in-volved the left foot. Thirteen patients sustained a medialdislocation and 5 sustained a lateral dislocation. The injurywas closed in 12 cases and open in 6; 2 of 13 (15%) medialdislocations were open, as were 4 of 5 (80%) lateral dislo-cations. The mechanism of injury was a motor-vehicle ac-cident (MVA) in 10 cases, sports-related trauma in 4 cases,and an accidental fall in the remaining 4 cases. Sevenpatients (3 with lateral and 4 with medial dislocations)reported associated fractures (2 multifragment fractures ofthe talus, 1 patient with fracture of the fourth and fifthmetatarsals, 1 with fracture of the calcaneus, 1 with fractureof the cuboid, and 2 with fractures of the posterior processof the talus). In 1 case of lateral dislocation, there was anassociated disruption of the posterior tibial artery and injuryto the saphenous vein.

The 6 open dislocations received emergency surgicaltreatment with meticulous wound and soft-tissue debride-ment, and, after reduction of the dislocation, reconstructionof associated fractures was performed in 4 cases. In 4 ofthese 6 patients, temporary fixation with Kirschner wireswas needed. Systemic antibiotics were administered. A bi-valved long-leg cast was then applied and replaced with ashort-leg cast after 48 hours. The short-leg cast was main-tained for 4 to 6 weeks, depending on the rate of wound andsoft-tissue healing and the presence of any associated frac-tures. None of the 6 patients with an open dislocationdeveloped an acute or chronic infection.

The closed dislocations were reduced under general an-esthesia in the emergency department; none of these casesrequired open reduction. The patients were immobilizedfirst in a long-leg cast followed by a short leg cast, after 48hours, for a period of 4 weeks (or 6 weeks in the presenceof a fracture). Patients were allowed immediate weightbear-ing as tolerated. Between the time of final treatment and thefollow-up evaluation, no patient required surgical interven-tion.

The average range of subtalar motion on the unaffected
side was 36.7 � 7°. On basis of the previously stated
VOLUME 43, NUMBER 3, MAY/JUNE 2004 167

, righ

criteria, 10 patients (56%) were classified as excellent (Fig1); each of these cases suffered a medial dislocation. Four ofthese patients sustained their dislocation during sport activ-ities, 3 were caused by an accidental fall, and the remaining3 by MVA. One of these patients also exhibited a fracture ofposterior process of the talus. Three of these 10 patientsshowed minimal radiographic signs of subtalar degenerativearthritis but were asymptomatic.

Three patients (17%) were classified as fair, including 1open medial, 1 closed medial, and 1 closed lateral disloca-tion. One patient sustained the injury during an accidentalfall and the other 2 were involved in an MVA. Two of thesepatients displayed associated fractures: a cuboid fracture inthe case of closed medial dislocation and a fracture of theposterior process of the talus associated with the closedlateral dislocation. The average of subtalar joint motion inthese patients was 21.3 � 3°. Radiographic signs of degen-erative arthritis involved the subtalar joint in 2 cases andboth the subtalar and the talonavicular joints in 1 case.

The remaining 5 cases (28%) showed poor results (Fig 2).All were open injuries and all were sustained during anMVA. Three of these 5 cases were lateral dislocations.

TABLE 1 Data of 18 patients with subtalar dislocation

Case Sex Age(yr)

Side Cause Type Open/Closed

AssociatedLesion

Treatmen

1 M 27 L MVA Medial Open — Surgica

2 M 36 R Sports Medial Closed — CR3 M 40 L MVA Lateral Open Fracture of

talusSurgica

4 F 59 L Fall Medial Closed — CR5 M 22 R MVA Lateral Closed Fracture of

posteriorprocesstalus

6 F 49 R Sports Medial Closed — CR7 F 18 R Fall Medial Closed — CR8 M 21 L MVA Lateral Open Fracture of

talusSurgica

9 F 42 R Sports Medial Closed — CR10 M 27 L Sports Medial Closed — CR11 F 38 L MVA Lateral Open Fracture of

4th, 5thmetatarsal

Surgica

12 F 46 R MVA Medial Closed — CR13 F 31 R MVA Medial Open Fracture of

calcaneumSurgica

14 M 20 L MVA Medial Closed — CR15 M 33 R Fall Medial Closed Fracture of

cuboidCR

16 M 43 L Fall Medial Closed — CR17 M 38 R MVA Lateral Open Rupture of

posteriortibialartery

Surgica

18 M 25 R MVA Medial Closed Fracture ofposteriorprocesstalus

CR

Abbreviations: CR, closed reduction; F, female; L, left; M, male; R

Complex concomitant fractures were found in 4 of these


patients. The mean subtalar range of motion in this group ofpatients was 7.2 � 2° at follow-up. Two of 5 patientsrequired an ankle brace with rigid stays for comfort. De-generative arthritis involving both the subtalar and the tal-onavicular joints was found in all 5 cases. There were nocases of talar avascular necrosis.

Discussion

Previous reports of peritalar dislocation have found me-dial dislocation to predominate (3, 8–10, 12, 14). Ourfindings agree; we found 73% of our cases to be medialdislocations. Most peritalar dislocations can be treated withclosed reduction. As long as reduction is obtained quickly,skin loss caused by pressure necrosis from the underlyingdislocation can be prevented (4). Generally, reduction pro-vides a stable result because of the distinctive shape of thearticular surfaces. Surgical intervention should be per-formed in open injuries or soon after the failure of closedreduction (10).

The need for open reduction has been reported to range

mobilization(wk)

Sequelae Result

Pain Mobility SubtalarArthritis

TalonavicularArthritis

4 Slight Slightlyimpaired

Moderate Moderate Fair

4 None Total — — Excellent6 Considerable Greatly

impairedSerious Serious Poor

4 None Total — — Excellent4 Slight Slightly

impairedModerate — Fair

4 None Total — — Excellent5 None Total Slight — Excellent6 Considerable Greatly


4 None Total — — Excellent4 None Total — — Excellent4 Considerable Greatly


5 None Total — — Excellent6 Considerable Greatly


4 None Total — — Excellent6 Slight Slightly

impairedModerate — Fair

5 None Total Slight — Excellent4 Considerable Greatly


4 None Total Slight — Excellent

t.

t Im

l

l

l

l

l

l

from 2% to 38% (3, 15). Our series found no cases of

FIGURE 1 Radiographic viewsshowing medial peritalar dislo-cation caused by MVA. Notedislocation of talonavicular andsubtalar joints. (A) Anteroposte-rior view. (B) Lateral view. (C)Lateral radiograph 7 years afterreduction, showing mild signsof midtarsal joint degenerativearthritis.


FIGURE 2 Radiographs oflateral peritalar dislocation. (A)Anteroposterior view. (B) Lat-eral view. (C) Lateral radio-graph 19 years after reduc-tion, showing severe signs ofankle and subtalar joint de-
generative arthritis.

irreducibility after gentle closed attempts at reduction.However, it should be noted that multiple attempts are notrecommended because this can compromise the surroundingneurovascular tissues, causing soft-tissue necrosis. None ofour patients developed this complication. The most com-mon reported cause of irreducibility for medial dislocationsis buttonholing of the talar head through the extensor reti-naculum or extensor digitorum brevis (16–18). This resultsin entrapment of the talar head and neck or impingement ofthe talus and calcaneus with fractures of the articular sur-faces (5). Heck (19) described the impingement of the deepperoneal neurovascular bundle as a rare cause of irreduciblemedial dislocation. The most frequent cause of irreduciblelateral dislocations is superolateral displacement of the tib-ialis posterior or the flexor digitorum longus tendons ontothe lateral neck of the talus (20).

The results reported in this series may appear to be quiteunsatisfactory because almost half of our patients experi-enced a fair or poor outcome. All 10 patients who exhibitedan excellent result sustained the injury during sports, anaccidental fall, or a low-energy MVA. All were closedmedial dislocations and in only one case was an associatedfracture reported.

All patients who sustained high-energy injuries showed apoorer prognosis. Previous studies have also reported thatopen injuries and concomitant intraarticular subtalar or tal-onavicular fractures contribute to poor results (3, 10, 14,15). These factors may be independent risk factors for thedevelopment of posttraumatic arthrosis (3).

Stiffness of the hindfoot is the most common impairmentafter subtalar dislocation (5). Many patients in other serieshave shown a loss of subtalar motion, especially after lateraldislocation (3, 10, 12, 21, 22). In our experience, 45% ofpatients exhibited a reduction of subtalar motion, with aresidual range of motion of 6° to 24°. Again, the majority ofthese cases were associated with high-energy trauma, con-comitant fractures, or open injuries. Two patients with poorresults wore an ankle brace with rigid stays to alleviate thepain related to degenerative arthritis.

Hindfoot degenerative arthritis after subtalar dislocationis the most common cause of long-term pain and disability(3, 8, 14, 23). Other studies have attributed the extent ofdegenerative arthritis to the presence of lateral dislocation,the force of the injury, the amount of soft-tissue injury, andthe presence of concomitant hindfoot fractures (3, 14, 21).More favorable long-term results are reported in cases inwhich only low-energy subtalar dislocations were evaluated(24). A mild degree of subtalar joint degenerative arthritishas also been reported to be well tolerated (5). Our 3patients with mild radiographic evidence of degenerativearthritis likewise were completely asymptomatic andshowed an excellent result.

In our series, we noted severe signs of subtalar and
talonavicular articular damage in 5 patients (28%). In the 11
patients with some degree of radiographic degenerativechanges, the subtalar joint was always involved. The tal-onavicular joint was also involved in 5 of these 11 patients,particularly in cases with fair or poor results. We believethat the talonavicular joint is probably involved secondarilybecause of the functional interrelationship of the subtalarand midtarsal joints. In fact, the mechanism of peritalardislocation dictates that the talonavicular joint dislocatesfirst, suggesting that the force required to dislocate this jointis not as significant as for the subtalar joint. Therefore,involvement of the subtalar joint may indicate a high-energyinjury, with a disruption of subtalar motion, as its long-termconsequence. The interrelationship between the subtalar andmidtarsal joints may then lead to secondary degenerativearthritis of talonavicular joint.

Degenerative arthritis may also occur without an associ-ated fracture (14). Our study included patients with fair orpoor results without an associated articular fracture, a find-ing reported by other authors (14). The subtalar joint isprone to arthrosis after isolated peritalar dislocation, partic-ularly in high-energy trauma, because the calcaneus slidespast the talus during dislocation, causing compressive andshearing forces that may result in injury to the cartilage (5).We believe that the high forces required to cause peritalardislocation can create cartilagineus lesions that are notdetected on radiographs but affect the patient’s clinicaloutcome. It may be interesting to perform imaging, such ascomputed tomography, of hindfoot routinely to evaluate thearticular surface after peritalar dislocation. Computed to-mography has been used to detect an osteochondral fractureof head of talus and to visualize periarticular anatomy (9,18, 24). In one study, postreduction tomography was rec-ommended because plain radiographs had failed to diagnosethe osteochondral fracture (3). Moreover, the authors foundopen reduction and fixation or excision of the fragment toreduce the risk of poor results (3). None of our patientsunderwent fusion. Surgery was proposed to 2 patients whopresented with poor results, but they refused.

Another possible complication of peritalar dislocation ispostreduction instability. Zimmer and Johnson (8) found 5of their 8 cases of peritalar dislocation experienced insta-bility and attributed the high incidence to an inadequateperiod of immobilization. For this reason, they recom-mended immobilizing patients in a plaster cast for morethan 4 weeks. On the other hand, prolonged immobilizationcan cause articular stiffness with subsequent loss of func-tion. Christensen et al (14) immobilized patients in a longcast for 8 weeks after reduction of dislocation and foundthat 19 of their 30 patients developed subtalar joint arthro-sis. Buckingam (9) reported that reduction of subtalar mo-tion was associated with 6 weeks of immobilization. De Lee(3) found that casting for 3 weeks gave better outcomes. Weconsider a period between 4 and 6 weeks an acceptable
compromise. In our series, no patient developed a recurrent

dislocation or instability of the subtalar joint. Conversely,patients treated with 6 weeks of plaster immobilizationexhibited less favorable functional and radiographic results,although it is unclear whether this is related to the length ofcast immobilization or the type of lesion. Talar necrosisafter peritalar dislocation is quite rare despite of this bone’ssusceptibility. No case was reported in our series at amaximum follow-up of 26 years, and few have been re-ported (3, 14).

The principal limitation of this study is the absence of astandardized rating scale. However, the long-term follow-upand the blinded evaluation of our clinical and radiographicresults allowed us to delineate the characteristics of peritalardislocations and the factors that contribute to their progno-sis.

Conclusion

Our study confirms that good results can be obtained inthe treatment of medial, low-energy peritalar dislocations.Early reduction is essential to prevent skin loss caused bypressure necrosis from the underlying dislocation; in nocase was open reduction required for a closed injury in ourseries. Degenerative arthritis is common after peritalar dis-location, particularly in cases of high-energy mechanisms,lateral dislocations, and dislocations associated with periar-ticular fractures. This complication is the single most im-portant cause of long-term pain and disability, but may bewell tolerated by some patients. Surgeons should be awareof the predictors of fair or poor outcome and counsel pa-tients accordingly. Secondary hindfoot instability after thisinjury was not observed in our series, and we believe thatthe time of cast weightbearing immobilization was judiciousand contributed to the low incidence of this complication.

References

1. Freund KG. Subtalar dislocations: a review of the literature. J FootSurg 28:429–432, 1989.

2. Bohay DR, Manoli A II. Subtalar joint dislocations. Foot Ankle Int12:803–808, 1995.

3. DeLee JC, Curtis R. Subtalar dislocation of the foot. J Bone Joint Surg64A:433–437, 1982.


4. Broca P. Memoire sur les luxations sous-astragaliennes. Mem Soc Chir3:566–656, 1853.

5. Saltzman C, Marsh JL. Hindfoot dislocations: when are they notbenign? J Am Acad Orthop Surg 5:192–198, 1997.

6. Grantham SA. Medial subtalar dislocations: five cases with a commonetiology. J Trauma 4:845–849, 1964.

7. Kanda T, Sakai H, Tamai K, Takeyama N, Saotome K. Anteriordislocation of the subtalar joint: a case report. Foot Ankle Int 22:609–611, 2001.

8. Zimmer TJ, Johnson KA. Subtalar dislocations. Clin Orthop 238:190–194, 1989.

9. Buckingham WW Jr. Subtalar dislocation of the foot. J Trauma 13:753–765, 1973.

10. Merianos P, Papaglannakos K, Hazis A, Tsafantakis E. Peritalar dis-locations: a follow-up report of 21 cases. Injury 19:439–442, 1988.

11. Kleiger B, Ahmed M. Injuries of the talus and its joints. Clin Orthop121:243–262, 1976.

12. Monson ST, Ryan JR. Subtalar dislocation. J Bone Joint Surg 63A:1156–1158, 1981.

13. Inman VT, Mann RA. Principles of examination of the foot and ankle.In Du Vries’ Surgery of Foot, ed 4, pp 22–42, edited by RA Mann,Mosby, St. Louis, 1978.

14. Christensen SB, Lorentzen JE, Krogjoe O, Sneppen O. Subtalar dis-locations. Acta Orthop Scand 48:707–711, 1977.

15. Rodriguez-Merchan EC. Subtalar dislocations: a study of 19 cases. IntOrthop 21:142–145, 1997.

16. Cohen MG, Garcia JF, Worrel RV. Imaging rounds. Orthop Rev20:466–472, 1991.

17. Leitner B. Obstacles to reduction in subtalar dislocations. J Bone JointSurg 36A:299–306, 1954.

18. Pehlivan O, Akmaz I, Solakoglu C, Rodop O. Medial peritalar dislo-cation. Arch Orthop Trauma Surg 122:541–543, 2002.

19. Heck BE, Ebraheim NA, Jackson WT. Anatomical considerations ofirreducible medial subtalar dislocation. Foot Ankle Int 17:103–106,1996.

20. Waldrop J, Ebraheim NA, Shapiro P, Jackson WT. Anatomical con-siderations of posterior tibialis tendon entrapment in irreducible lateralsubtalar dislocation. Foot Ankle 13:458–461, 1992.

21. Heppenstall RB, Farahvar H, Balderston R, Lotke P. Evaluationand management of subtalar dislocations. J Trauma 20:494–497,1980.

22. Tucker DJ, Burian G, Boylan JP. Lateral subtalar dislocation: reviewof the literature and case presentation. J Foot Ankle Surg 37:239–247,1998.

23. Ruiz Valdivieso T, de Miguel Vielba JA, Hernandez Garcia C, Cas-trillo AV, Alvarez Posadas JI, Sanchez Martin MM. Subtalar disloca-tion: a study of nineteen cases. Int Orthop 20:83–86, 1996.

24. Perugia D, Basile A, Massoni C, Gumina S, Rossi F, Ferretti A.Conservative treatment of subtalar dislocations. Int Orthop 26:56–60,2002.

Peritalar dislocations: a retrospective study of 18 cases

Documents

Transcript of Peritalar dislocations: a retrospective study of 18 cases