Crossover Second Toe: Demographics, Etiology, and Radiographic Assessment

Transcription

1 FOOT &ANKLE INTERNATIONAL Copyright 2007 by the American Orthopaedic Foot & Ankle Society, Inc. DOI: /FAI Crossover Second Toe: Demographics, Etiology, and Radiographic Assessment Ari J. Kaz, M.D.; Michael J. Coughlin, M.D. Boise, ID ABSTRACT INTRODUCTION Background: The purpose of this study was to determine the demographics, etiology, and radiographic findings associated with a crossover second toe deformity. Methods: Patients treated operatively for a crossover second toe deformity between 2001 and 2006 were identified. Charts were reviewed for clinical information, and radiographs were examined for pertinent angular measurements. Results: Of 169 patients in the study, 146 (86%) were women. The mean age at surgery was 59 (range 33 to 87) years. The most common complaints of preoperative pain were at the second (156 patients) and first (35 patients) metatarsophalangeal joints (several patients had more than one area of pain). A positive drawer sign was noted in 112 patients. The mean second and third metatarsophalangeal joint angles were 3 degrees and 6 degrees, respectively. There was a significant association of hallux valgus with first metatarsophalangeal joint arthritis (p < 0.01). The relative length of the second metatarsal averaged 0.2 mm less than the first metatarsal. Conclusions: Crossover second toe deformity had a peak incidence in women over the age of 50 years. There was an increased incidence of both hallux valgus and first metatarsophalangeal joint degenerative arthritis in the patient cohort. A positive drawer sign was a reliable and consistent physical examination finding. The most reliable radiographic indicator of a second crossover toe was medial angulardeviation of the second metatarsophalangeal joint in relationship to the third metatarsophalangeal joint angle, although the angle was not necessarily a negative value. There was no correlation between a crossover second toe deformity and second metatarsal length, medial cortex thickness or shaft thickness, the 1 2 intermetatarsal angle, metatarsus adductus, metatarsus primus elevatus, or pes planus. Key Words: Crossover Toe; Pes Planus; Metatarsus Primus Elevatus; Second Ray Overload Corresponding Author: Michael J. Coughlin, M.D. Director, Idaho Foot and Ankle Fellowship 901 N. Curtis Road, Suite 503 Boise, ID footmd@aol.com For information on prices and availability of reprints, call X Pain, inflammation, and subluxation of the second metatarsophalangeal joint are common forefoot problems. 3,7,13,25,45 Frequently the second toe deviates medially or dorsomedially in reference to both the hallux and third toe, leading to what has been termed a crossover toe deformity. 5,7 9 This deformity often begins with pain at the second metatarsophalangeal joint, followed by a subtle medial inclination at the metatarsophalangeal joint manifested as a slight increase in the interval between the second and third toes. 5,7 9,40 As the deformity progresses, the proximal phalanx hyperextends and can deviate dorsomedially. 5,6,10 It may progress to the point where the second toe crosses up and over the hallux, coming to lie on the dorsal surface of the great toe (Figure 1). 5 Factors contributing to this deformity are thought to include extrinsic causes such as trauma, 39,42 high-fashion footwear, 3,5,10,15,20,27,43,46 pressure on the second toe from a hallux valgus deformity, 3,5,6,19,13,31 or impingement from the neighboring third toe. 5,13 Anatomic factors such as an anomalous muscle slip 29 or muscular imbalance between the intrinsic and extrinsic musculature of the toes with resultant hyperextension of the proximal phalanx also have been postulated to be a cause of metatarsophalangeal joint instability. 1,3,19,27,43 The frequency of occurrence at the second metatarsophalangeal joint has led some authors to postulate that it is caused by a long second metatarsal. 3,5,9,13,19,27,32,45,48 Although first ray hypermobility and forefoot pronation with hindfoot valgus have been implicated as causes of second metatarsal overload with subsequent second metatarsophalangeal joint deformity, 27,34 no evidence has been presented to substantiate these notions. While the progression of the deformity and its pathophysiology have been well documented in the literature, the demographics, etiology, and radiographic findings associated with a crossover toe deformity have not been defined. MATERIALS AND METHODS A retrospective review of charts and radiographs from the

2 1224 KAZ AND COUGHLIN Foot & Ankle International/Vol. 28, No. 12/December 2007 A B C D Fig. 1: Clinical pictures of a crossover second toe. A, Dorsal view. B, Plantar view. CandD, Anteroposterior radiographs of the progression of a crossover second toe. senior author s practice (M.J.C.) was undertaken. The records of 326 consecutive patients with a diagnosis of a second crossover toe that had operative correction between January of 2001 and December of 2006 were reviewed. Approval from the institutional review board was obtained before commencement of the study. Exclusion criteria included previous ipsilateral midfoot or hindfoot surgery (17 patients), trauma (nine patients), inflammatory arthritis (nine patients), missing or inadequate radiographs (15 patients), clinically apparent lateral deviation (26 patients), or pure dorsal deviation (81 patients) at the second metatarsophalangeal joint. This left 169 patients who were eligible for the study) of whom 146 (86%) were women and 23 (14%) were men. The mean age at surgery was 59 (range 33 to 87) years. The right foot was involved in 89 patients, and the left in 80 patients. The duration of symptoms ranged from 3 weeks to 40 years (mean 31 months). All radiographic measurements were made by the junior author (A.J.K.), and all chart reviews were done by the senior author (M.J.C.). Charts were reviewed for clinical information, including patient demographics, chief preoperative and other preoperative complaints, duration and location of preoperative symptoms, description of forefoot deformity, presence of a drawer sign, 45 and a 10-point preoperative pain score. Standing weightbearing anteroposterior and lateral radiographs were obtained using a standardized technique, with the radiographic beam located 40 inches from the subject s foot and the cassette placed directly adjacent to the foot. All radiographic measurements were made with a hinged goniometer with 5-degree increments and an electronic digital slide caliper (catalog # 721 A- 6/150; L.S. Starett, Athol, Massachusetts). The caliper has

3 Foot & Ankle International/Vol. 28, No. 12/December 2007 CROSSOVER SECOND TOE 1225 a measurement error of 0.03 mm. Preoperative weightbearing radiographs were examined for angular measurements for the following: the first, second, and third metatarsophalangeal joint angles; 11 the first, second, and third intermetatarsal angles 11 (Appendix 1); elevatus of the first metatarsal 30 (Appendix 2); relative length of the first, second, and third metatarsals 28 (Appendix 3); metatarsus adductus 14 (Appendix 4); the Meary angle (the lateral talometatarsal angle); 22,26 the lateral talocalcaneal angle 44 (Appendix 5); the anteroposterior talonavicular coverage angle 21 (Appendix 6); the second metatarsal shaft length, intramedullary canal thickness, medial cortex thickness, and shaft thickness 24,41 (Appendix 7); and the presence of impingement between the second and third metatarsals. The presence or absence of first metatarsophalangeal degenerative joint disease was recorded. Arthritis was graded as either absent or present. (Using the criteria of Coughlin and Shurnas, 12 absent was grade 0 and; present was from grade 1 through 4). In assessing radiographs for hallux valgus, a first metatarsophalangeal joint angle of 16 degrees or more was considered a bunion deformity. 16 Using the axis of the third metatarsal joint articulation as a means of comparison, the second metatarsophalangeal joint angle was considered abnormal if it was 8 degrees or less. 9 All second metatarsophalangeal joints with a negative angle (medial deviation) were included; also, second rays with a metatarsophalangeal joint angle between 0 and 8 degrees were included if there was a minimum of 5 degrees difference between the second and third metatarsophalangeal joint angles, the second being inclined more medial than the third. Statistical Analysis Statistical analysis was performed using SPSS 11.5 software (Chicago, Illinois), with an alpha value set at p < 0.05 for all analyses. The Pearson correlation matrix, Spearman s rho, independent sample t-tests, and Fisher exact tests were used to evaluate the data. Positive coefficients closer to 1 indicate strong correlation, and values closer to 0 indicate weak or no correlation. RESULTS Symptoms The most common location of preoperative pain was the second metatarsophalangeal joint (156), followed by the first metatarsophalangeal joint (35) and the third metatarsophalangeal joint (28). (All areas of pain and subjective complaints were recorded, and several subjects had more than one area of pain or subjective complaints.) The preoperative 10-point visual analogue pain scale averaged 6.5 points (range 1 to 10). The most common subjective complaint was second metatarsophalangeal joint malalignment (166), followed by hallux valgus (69). Other complaints consisted of symptoms of hallux rigidus (16), an interdigital neuroma (13), and hallux varus (5). Five patients had prior interdigital neuroma surgery without relief of symptoms. Physical Examination Findings One hundred and sixty-four patients (97%) demonstrated dorsomedial deviation of the second toe. The remaining five joints had eventually dislocated and had assumed a complex alignment that could not be characterized as pure dorsomedial deviation. A positive drawer sign was documented in 112 patients. One had a documented negative drawer sign, and 51 had no record in the chart of a drawer test being performed. However, a drawer test was not performed on patients who had severe fixed second toe deformities or those with obvious painful dorsomedial deviation of the second toe. A second hammertoe was present in 75 patients (44%), and a mallet toe in nine. Twelve of 169 patients (7%) developed a callus or an intractable plantar keratosis beneath the second metatarsal head. Twenty-five patients had documented palpable thickness, swelling, or both about the second metatarsophalangeal joint. Clinically, hallux valgus was present in 83 patients, hallux rigidus in 24, hallux varus in 11, and hallux valgus interphalangeus in one. Third toe deformities included a hammertoe in 31 patients, with five patients presenting with dorsomedial deviation, one of whom had a positive drawer. Six patients had a third mallet toe, and four developed an intractable plantar keratosis or callus beneath the third metatarsal head. Radiographic Findings Hallux valgus and lesser metatarsal angular measurements (Table 1) Of the 169 patients, 83 (49%) had a hallux valgus angle of more than or equal to 16 degrees. 16 The average hallux valgus angle for the entire cohort was 15.5 degrees (range 36 to 43 degrees); the average 1 2 intermetatarsal angle was 12 degrees (range 1 to 20 degrees). The association between the hallux valgus angle and the 1 2 intermetatarsal angle was highly significant (r = 0.749, p < 0.01). One hundred and nine of 169 (65%) patients had a negative second metatarsophalangeal joint angle, and the remaining 60 patients had a second metatarsophalangeal joint angle between 0 and 8 degrees, with a minimum of 5 degrees difference between their respective second and third metatarsophalangeal joint angles and an obvious spread between the two toes. The average second metatarsophalangeal joint angle was 3 degrees (range 30 to 8 degrees); the average 2 3 intermetatarsal angle was 2.3 degrees (range 2 to7 degrees). There was a statistically significant but low correlation between the second metatarsophalangeal joint angle and both the hallux valgus angle (r = 0.445, p < 0.01), and the 2 3 intermetatarsal angle (r = 0.212, p < 0.01). There was no correlation between the second metatarsophalangeal joint angle and the 1 2 intermetatarsal angle. The average third metatarsophalangeal joint angle was 6 degrees (range 15 to 23 degrees). The average 3 4

4 1226 KAZ AND COUGHLIN Foot & Ankle International/Vol. 28, No. 12/December 2007 Table 1: Statistical correlation strength (r and p values) for angular measurements (values in bold denote statistical significance) HVA 2 nd MTPA 3 rd MTPA 1,2 IMA 2,3 IMA 3,4 IMA HVA low very low high r = very low x (r = 0.445) (r = 0.186) (r = 0.749) p = (r = 0.165) 2 nd MTPA low moderate r = very low low (r = 0.445) x (r = 0.513) p = (r = 0.212) (r = 0.288) 3 rd MTPA very low moderate r = very low r = (r = 0.186) (r = 0.513) x p = (r = 0.187) p = ,2 IMA high r = r = r = r = (r = 0.749) p = p = x p = p = ,3 IMA r = very low very low r = very low p = (r = 0.212) (r = 0.187) p = x (r = 0.156) 3,4 IMA very low very low r = r = very low (r = 0.165) (r = 0.288) p = p = (r = 0.156) x MPE r = r = r = r = r = r = p = p = p = p = p = p = nd MT low low r = very low r = very low protrusion distance (r = 0.423) (r = 0.299) p = (r = 0.244) p = (r = 0.234) 3 rd MT low low low very low r = very low protrusion distance (r = 0.363) (r = 0.280) (r = 0.166) (r = 0.186) p = (r = 0.228) Metatarsus r = r = r = r = very low very low adductus p = p = p = p = (r = 0.237) (r = 0.204) Meary r = very low very low r = r = very low angle p = (r = 0.219) (r = 0.223) p = p = (r = 0.275) Lateral TC r = r = very low r = r = r = Angle p = p = (r = 0.167) p = p = p = AP TN r = r = r = r = r = very low coverage p = p = p = p = p = (r = 0.161) 2 nd IMT r = r = r = r = r = r = p = p = p = p = p = p = nd ST r = r = r = r = r = r = p = p = p = p = p = p = nd MCT r = r = r = r = r = r = p = p = p = p = p = p = nd shaft r = r = r = low r = r = length p = p = p = (r = 0.228) p = p = HVA, hallux valgus angle; MTPA, metatarsophalangeal joint angle; IMA; intermetatarsal angle; MPE, metatarsus primus elevatus; MT, metatarsal; TC, talocalcaneal; AP, anteroposterior, TN, talonavicular; IMT, intramedullary canal thickness; ST, shaft thickness; MCT, medial cortical thickness.

5 Foot & Ankle International/Vol. 28, No. 12/December 2007 CROSSOVER SECOND TOE 1227 intermetatarsal angle was 6 degrees (range 2 to 12 degrees). A statistically significant moderate correlation was found between the second metatarsophalangeal joint angle and the third metatarsophalangeal joint angle (r = 0.513, p < 0.01). We examined the relationship between the 83 patients with a hallux valgus angle of 16 degrees or more (the bunion group ) and their corresponding second metatarsophalangeal joint angle, as well as the 86 patients without a bunion and their corresponding second metatarsophalangeal joint angle. The bunion group had an average second metatarsophalangeal joint angle of 0.6 degrees, while the remaining subjects without a bunion had an average second metatarsophalangeal joint angle of 5.3 degrees, a difference that was statistically significant (z = 4.7, p < 0.01). Hallux Valgus, First MTP DJD, and the Second MTP Angle While 24 patients had a diagnosis of hallux rigidus, another 24 patients (a total of 48 patients) had radiographic evidence of first metatarsophalangeal degenerative joint disease, including 16 of the 83 patients (19%) with a bunion and 32 of the 86 (37%) without a bunion. The association between increasing angular deformity of the first metatarsophalangeal joint and first metatarsophalangeal joint arthritis was statistically significant (χ 2 = 3.1, p < 0.01). Ten of 60 patients (17%) with a positive second metatarsophalangeal joint angle had evidence of first metatarsophalangeal degenerative joint disease while 37 of 109 (34%) with a negative second metatarsophalangeal joint angle had evidence of first metatarsophalangeal degenerative joint disease. One hundred and twenty-one subjects (72%) had no evidence of first metatarsophalangeal degenerative joint arthritis. (One patient had a prior first metatarsophalangeal joint fusion and was therefore not included.) The association between the medial deviation of the second metatarsophalangeal joint angle and first metatarsophalangeal degenerative joint disease was statistically significant (χ 2 = 7.5, p < 0.01). The mean second metatarsophalangeal angle for the 121 patients who did not have first metatarsophalangeal degenerative joint disease was 2.2 degrees. For those 48 patients with first metatarsophalangeal degenerative joint disease the mean second metatarsophalangeal angle was 4.9 degrees. This difference was statistically significant (z = 2.7, p<0.01). Measurements of Second Metatarsal Length and Evidence of Overload (Table 2) Using Hardy s arc method of measurement 28 (Figure 4) for metatarsal length, the relative length of the second metatarsal averaged 0.2 mm less than the first metatarsal, and the relative length of the third metatarsal averaged 3.8 mm less than the first metatarsal. Seventy-four patients (44%) had a second metatarsal that was longer than the first. The average second metatarsal shaft thickness was 8.5 mm, the average intramedullary canal thickness was 3.5 mm, and the average medial cortical thickness was 3.4 mm (Figure 8). 41 A statistically significant, low-negative correlation was found between the second metatarsal protrusion distance and both the second metatarsophalangeal joint angle (r = 0.299, p < 0.01) and the hallux valgus angle (r = 0.423, p < 0.01). The relationships between the second metatarsophalangeal joint angle and the second medial cortical thickness (p = 0.573) and second shaft length (p = 0.554) were not statistically significant. A statistically significant highpositive correlation was found between the second and third metatarsal protrusion distances (r = 0.850, p < 0.01). Pes Planus (Table 3) The average Meary angle (the lateral talometatarsal angle) for the entire cohort was 3.3 degrees (range 15 to 40; normal is 4 to 4 degrees. 22,44 ). A very low, but statistically significant, negative relationship was found between the second metatarsophalangeal joint angle and the Meary angle (r = 0.219, p < 0.01). The mean second metatarsophalangeal joint angle for persons with a Meary angle of less than 4 degrees (thus considered pes planus 22 ) was 0.4 degrees. For those with a Meary angle of 4 degrees or more (which represents the foot with a normal arch and those considered to have a cavus posture), the mean second metatarsophalangeal joint angle was 3.7 degrees. The difference between groups was statistically significant (z = 2.6, p < 0.01). The average anteroposterior talonavicular coverage angle was 16.8 degrees (range 0 to 37; normal is 14 ±3 degrees 33 ). The average lateral talocalcaneal angle was 45.7 degrees (range 2 to 63; normal is 25 to 50 degrees 44 ). No statistically significant relationship was found between the second metatarsophalangeal joint angle and either the lateral talocalcaneal angle (r = 0.067, p = 0.387) or the anteroposterior talonavicular coverage angle (r = 0.141, p = 0.067). Impingement Eighty patients had radiographic evidence of impingement. This was defined as less than 1 mm of separation between the second and third metatarsal heads on the anteroposterior weightbearing radiograph. Ten of the patients enrolled during the last 3 months of the study were evaluated under fluoroscopy, and when a true anteroposterior radiograph of the foot was obtained none demonstrated impingement as previously defined (Figure 2). Metatarsus Adductus The average metatarsus adductus angle was 15.5 (range 3 to 31) degrees. A very low, but statistically significant, positive relationship was found between the metatarsus adductus angle and the 2 3 intermetatarsal angle (r = 0.237, p < 0.01), the 3 4 intermetatarsal angle (r = 0.204, p < 0.01), and second metatarsal medial cortical thickness (r = 0.159, p < 0.01). There was no correlation between the metatarsus

6 1228 KAZ AND COUGHLIN Foot & Ankle International/Vol. 28, No. 12/December 2007 Table 2: Statistical correlation strength (r and p values) of measurements of second metatarsal (values in bold denote statistical significance) MPE 2 nd MT protrusion distance 3 rd MT protrusion distance 2 nd IMT 2 nd ST 2 nd MCT 2 nd shaft length HVA r = low low r = r = r = r = p = (r = 0.423) (r = 0.363) p = p = p = p = nd MTPA r = low low r = r = r = r = p = (r = 0.299) (r = 0.280) p = p = p = p = rd MTPA r = r = very low r = r = r = r = p = p = (r = 0.166) p = p = p = p = ,2 IMA r = very low very low r = r = r = low p = (r = 0.244) (r = 0.186) p = p = p = (r = 0.260) 2,3 IMA r = r = r = r = r = r = r = p = p = p = p = p = p = p = ,4 IMA r = very low very low r = r = very low r = p = (r = 0.234) (r = 0.228) p = p = (r = 0.125) p = MPE low low r = r = r = r = x (r =.0311) (r = 0.252) p = p = p = p = nd MT low high r = r = r = low protrusion distance (r = 0.311) x (r = 0.850) p = p = p = (r = 0.258) 3 rd MT low high r = r = r = r = protrusion distance (r = 0.252) (r = 0.850) x p = p = p = p = (Continued)

7 Foot & Ankle International/Vol. 28, No. 12/December 2007 CROSSOVER SECOND TOE 1229 Table 2: (Continued) MPE 2 nd MT protrusion distance 3 rd MT protrusion distance 2 nd IMT 2 nd ST 2 nd MCT 2 nd shaft length Metatarsus r = low low r = r = very low r = adductus p = (r = 0.290) (r = 0.423) p = p = (r = 0.159) p = Meary r = r = r = r = r = r = r = angle p = p = p = p = p = p = p = Lateral TC r = r = r = r = r = r = very low angle p = p = p = p = p = p = (r = 0.164) AP TN r = r = r = r = r = r = r = coverage p = p = p = p = p = p = p = nd IMT r = r = r = low very low low p = p = p = x (r = 0.484) (r = 0.256) (r = 0.249) 2 nd ST r = r = r = low low low p = p = p = 0741 (r = 0.484) x (r = 0.406) (r = 0.420) 2 nd MCT r = r = r = very low low very low p = p = p = (r = 0.256) (r = 0.406) x (r = 0.229) 2 nd shaft r = low r = low low very low length p = (r = 0.258) p = (r = 0.249) (r = 0.420) (r = 0.229) x HVA, hallux valgus angle; MTPA, metatarsophalangeal joint angle; IMA; intermetatarsal angle; MPE, metatrsus primus elevatus; MT, metatarsal; TC, talocalcaneal; AP, anteroposterior, TN, talonavicular; IMT, intramedullary canal thickness; ST, shaft thickness; MCT, medial cortical thickness.

8 1230 KAZ AND COUGHLIN Foot & Ankle International/Vol. 28, No. 12/December 2007 Table 3: Statistical correlation strength (r and p values) of measurements of pes planus (values in bold denote statistical significance) Metatarsus adductus Meary s angle Lateral TC angle AP TN coverage HVA r = r = r = r = p = p = p = p = nd MTPA r = very low r = r = p = (r = 0.219) p = p = rd MTPA r = very low very low r = p = (r = 0.223) (r = 0.167) p = ,2 IMA r = r = r = r = p = p = p = p = ,3 IMA very low r = r = r = (r = 0.237) p = p = p = ,4 IMA very low very low r = very low (r = 0.204) (r = 0.275) p = (r = 0.161) MPE r = r = r = r = p = p = p = p = nd MT low r = r = r = protrusion distance (r = 0.290) p = p = p = rd MT low r = r = r = protrusion distance (r = 0.423) p = p = p = Metatarsus r = r = r = adductus x p = p = p = Meary r = low low angle p = x (r = 0.422) (r = 0.378) Lateral TC r = low very low angle p = (r = 0.422) x (r = 0.250) AP TN r = low very low coverage p = (r = 0.378) (r = 0.250) x 2nd IMT r = r = r = r = p = p = p = p = nd ST r = r = r = r = p = p = p = p = nd MCT very low r = r = r = (r = 0.159) p = p = p = nd shaft r = r = very low r = length p = p = (r = 0.164) p = HVA, hallux valgus angle; MTPA, metatarsophalangeal joint angle; IMA; intermetatarsal angle; MPE, metatarsus primus elevatus; MT, metatarsal; TC, talocalcaneal; AP, anteroposterior, TN, talonavicular; IMT, intramedullary canal thickness; ST, shaft thickness; MCT, medial cortical thickness.

9 Foot & Ankle International/Vol. 28, No. 12/December 2007 CROSSOVER SECOND TOE 1231 A B Fig. 2: A, Anteroposterior weightbearing radiograph with apparent impingement between the second and third metatarsals. B, On anteroposterior fluoroscopy picture of the same foot rotated to obtain a true anteroposterior view impingement is no longer visible. adductus angle and either the second metatarsophalangeal joint angle (r = 0.142, p = 0.066) or the third metatarsophalangeal joint angle (r = 0.048, p = 0.533). Metatarsus Primus Elevatus The average metatarsus primus elevatus measurement was 4.2 (range 2.32 to 10.8) mm. Only six patients (4%) had a metatarsus primus elevatus measurement of 8 mm or more. A statistically significant low, negative correlation was found between the metatarsus elevatus measurement and both the second metatarsal protrusion distance (r = 0.311, p < 0.01) and the third metatarsal protrusion distance (r = 0.252, p < 0.01). There was no correlation between the second metatarsophalangeal angle and hypermobility of the first ray as defined by a metatarsus primus elevatus measurement of greater than 8 mm (r = 0.062, p = 0.424). DISCUSSION History and Demographics The pattern of dorsomedial subluxation of the second toe was initially described by Coughlin. 5 A gap between the second and third toes has been the classic clinical sign of second metatarsophalangeal joint instability. 5,7,40 Of the 169 consecutive patients in this study, 86% were women. This is consistent with reports in the literature on deformity of the second metatarsophalangeal joint where a female preponderance has been noted. 5,10,15,20,46 While attenuation or rupture of the joint capsule has been speculated to develop from chronic hyperextension of the lesser metatarsophalangeal joints from the use of high fashion shoewear, 3,5,10,15,20,27,43,46 this study did not specifically study shoewear history. The mean age at surgery was 59 (range 33 to 87) years. While a second crossover toe can occur in younger patients, it more commonly presents in women over age 50 years. One hundred and forty-one of the 169 patients (83%) in the current study were at least 50 years of age. Often a misdiagnosis of a second interspace interdigital neuroma is made when, in fact, instability of the second metatarsophalangeal joint has occurred. In two earlier reports, 5,9 10% to 14% of patients had prior unsuccessful neuroma surgery. In the current series, five patients (3%) had unsuccessful neuroma surgery before their diagnosis of second metatarsophalangeal joint instability. The mean duration of symptoms was 31 months. It was rare for a patient to present early, and the median time from initial diagnosis to surgery was 12 months. We believe this information supports the more chronic nature of this deformity. Even though close to half of the patients had a clinically evident hallux valgus deformity, the primary location of pain in 92% of patients was reported to be the second metatarsophalangeal joint. With regards to deformity, 98% of patients reported a complaint of malalignment of the second metatarsophalangeal joint. We conclude that even in those patients with a hallux valgus deformity, the primary complaint is pain and deformity of the second metatarsophalangeal joint, and the bunion deformity is secondary.

10 1232 KAZ AND COUGHLIN Foot & Ankle International/Vol. 28, No. 12/December 2007 Physical Examination The primary restraint to metatarsophalangeal joint subluxation is the collateral ligament complex alone or in combination with the plantar plate. Attenuation of these structures leads to metatarsophalangeal joint instability, 1,2,5,8,13,17 20, 32,48 with the subsequent onset of pain and joint deformity. 6,7,10,15,23,27,31,35 In the current study, 164 patients presented with dorsomedial deviation of the second toe, with a documented positive drawer sign in 112 cases. The drawer sign is the first objective sign of metatarsophalangeal joint instability. 8 We have found the drawer test 9,45 to be the most reliable and reproducible physical examination finding to confirm second metatarsophalangeal joint instability. Coughlin 9 reported three dislocations in patients who had symptoms for 6, 12, and 24 months. While acute traumatic second metatarsophalangeal joint dislocation has been reported, 4,39,42 in the current study those presenting with frank dislocation reported a long-standing history of symptoms (1 year, five feet; 2 years, two feet; 3 years, one foot; over 3 years, two feet). Coughlin 9 reported that six of eight patients with symptoms over 1 year in duration developed a fixed hammertoe deformity. In the current study, a second hammertoe developed in 75 patients (44%), a finding consistent with previous reports. We believe the progression of deformity as evidenced by a hammertoe deformity and eventual metatarsophalangeal joint dislocation occurs with chronic symptoms. Hallux valgus occurred in 49% of patients, hallux rigidus in 14%, and hallux varus in 7%. Third toe deformities included a hammertoe in 31 patients, with five patients presenting with dorsomedial deviation at the metatarsophalangeal joint. Only one patient had a positive drawer test at the third metatarsophalangeal joint. Symptoms and deformity of the third metatarsophalangeal joint may occasionally be found in patients with a second crossover toe. We find it noteworthy that only 12 patients (7%) had a documented callus or intractable plantar keratosis beneath the second metatarsal head. We suggest that a callus or intractable plantar keratosis is not a consistent finding in second metatarsophalangeal joint deformity. Radiographs Many reports have implicated extrinsic pressure on the second metatarsophalangeal joint from a hallux valgus deformity as a cause of second metatarsophalangeal joint malalignment. 3,5,6,8,13,18,31 Prior series reported 27% to 28% of patients with concomitant hallux valgus and crossover second toe deformities, 5,9 compared to 49% in the present series. The patients in the current series with a hallux valgus deformity had an average second metatarsophalangeal angle of 0.5 degrees, while the patients without a bunion deformity had an average second metatarsophalangeal angle of 5.3 degrees. As to why those with a bunion deformity had a smaller second metatarsophalangeal joint angular deformity is unknown. We suggest that it may be due to a splinting effect from lateral pressure on the second toe by the hallux, or it may be explained by the fact that those with a bunion came to surgery sooner than those with an isolated crossover second toe deformity. Indeed, with the passage of time, those with an isolated second toe deformity may have progressed to greater deformity. The relationship between a second crossover toe and first metatarsophalangeal joint arthritis has not been previously reported. There was a threefold increase in patients with degenerative arthritis of the first metatarsophalangeal joint in those with a negative second metatarsophalangeal joint angle or more severe deformity. Whether radiographic evidence of first metatarsophalangeal joint arthritis predisposes to more severe angular deformity of the second metatarsophalangeal joint is unknown. In earlier reports on crossover toes, the mean medial angulation of the second metatarsophalangeal joint was reported as 9 degrees and 2.8 degrees. The mean value of 3 degrees in the current study concurs with those values. The average third metatarsophalangeal angle was 6.0 degrees (range 15 to 23 degrees), which concurs closely with the reported normal angulation of 8 to 12 degrees. 7 9 The axis of the third metatarsophalangeal joint has been used as a comparison for second metatarsophalangeal joint malalignment. We chose the lower value in the literature as the upper value of second metatarsophalangeal joint malalignment. We also required a 5-degree differential between the second and third metatarsophalangeal joint angles and clinically apparent divergence of the second and third toes to qualify as medial deviation of the second toe. This explains the inclusion of digits with up to +8 degrees of lateral inclination. While the upper limit of radiographic angular malalignment that defines a second crossover toe is unknown, we believe that our current definition is reasonable. However, we suggest further study of what constitutes normal alignment of the second metatarsophalangeal joint. Pressure from the adjacent third metatarsal or toe leading to an impingement phenomenon also has been implicated as leading to medial deviation at the second metatarsophalangeal joint. 5,13,15 Nearly half of the patients evaluated in this study had radiographic evidence of impingement as noted on the available anteroposterior radiographs. Ten patients enrolled during the last 3 months of the study were evaluated under fluoroscopy, and when a true anteroposterior radiograph of the foot was obtained, none demonstrated second or third ray impingement (Figure 2). We believe that the impingement as seen on anteroposterior radiographs is caused by the overlap of the second and third metatarsals. Our fluoroscopy evaluations consistently demonstrated that upon rotation of the foot to obtain a true anteroposterior view, this impingement was no longer present. We suggest that impingement is likely a radiographic artifact and that it is not a true cause of second crossover toe deformity. Further study of a much larger cohort will be necessary to confirm these findings.

11 Foot & Ankle International/Vol. 28, No. 12/December 2007 CROSSOVER SECOND TOE 1233 Many reports have theorized that a second crossover toe deformity is caused by a long second metatarsal. 3,5,7 9,13, 19,27,32,45,48 Morton 36 suggested that a foot with a long second metatarsal was prone to develop metatarsalgia. Coughlin 5,9 reported a long second metatarsal associated with a crossover second toe deformity in 80 to 90% of cases. These studies 5,9 used Morton s novel method of measurement of metatarsal length in which a transverse line is drawn from the first to the third distal metatarsal articular surfaces. The radiographs from the cohort of the senior author s original paper on crossover toe deformity 5 were again reviewed, and using the Hardy and Clapham 28 method of measuring second metatarsal length only three of 14 patients (21%) had a long second metatarsal, which on average was 0.5 mm shorter than the first metatarsal. In the current study, using Hardy and Clapham s 28 arc method of measurement, the relative length of the second metatarsal averaged 0.2 mm less than the first metatarsal, concurring very closely with the senior author s original cohort when using the same measurement technique. Seventy-four of 169 patients (44%) in our study had a second metatarsal that was longer than the first. Fewer than half of the feet in this study had a long second metatarsal and, hence, we find it difficult to support the notion that a long second metatarsal is a cause of this deformity. When planning any shortening osteotomy of the second metatarsal, we believe an accurate measurement of the true length of the first, second, and third metatarsals is essential, regardless of underlying angular deformity. Hypermobility of the first ray has been implicated as a cause of second metatarsal overload 27,34,36 38 with subsequent lateral metarsalgia. The normal value for metatarsus primus elevatus measured on a weightbearing lateral radiograph is 8 mm or less. 30 In the current study, the average metatarsus primus elevatus value was 4.2 mm (range 2.32 to 10.8 mm). Six patients had a metatarsus primus elevatus measurement of greater than 8 mm. There was no correlation between metatarsus primus elevatus and a crossover second toe deformity. Pes planus also has been implicated as a cause of second metatarsal overload. 27,34 The mean anteroposterior talonavicular coverage angle, the mean lateral talocalcaneal angle and the mean lateral talometatarsal angle (Meary angle) were all within normal limits. Furthermore, there was no statistically significant relationship between the second metatarsophalangeal joint angle and the lateral talometatarsal angle, the lateral talocalcaneal angle, or the anteroposterior talonavicular coverage angle. The average second metatarsophalangeal joint angle of those patients with flatfeet as defined by a Meary angle of less than 4 degrees was more positive (average of 0.4 degrees) than those with a normal or cavus arch as defined by a Meary angle greater than 4 degrees (average of 3.7 degrees), indicating a trend towards less angular deformity at the second metatarsophalangeal joint in those patients with a lower Meary angle or a flatfoot. While there was a very low negative correlation between the second metatarsophalangeal angle and Meary angle (r = 0.219, p < 0.01), the exceedingly low correlation leads us to question its significance. We conclude that there is little or no relationship between radiographic flatfoot measurements and the severity of medial deviation at the second metatarsophalangeal joint in patients with a crossover second toe deformity. The average metatarsus adductus angle was 15.6 (range 3 to 31) degrees, just above the normal value of 15 degrees. 6 There was no correlation between metatarsus adductus and the second metatarsophalangeal angle. Morton 36,37,38 and others 9 have suggested that a manifestation of first ray hypermobility and resultant pes planus is cortical hypertrophy of the second metatarsal. Grebing and Coughlin 24 evaluated the medial cortical thickness of the second metatarsal in patients with hallux valgus and reported a mean thickness of 3.1 mm. In the current study, the average medial cortical thickness was 3.4 mm, a value consistent with that previously reported by Grebing and Coughlin 24. We found no association between medial cortical hypertrophy of the second metatarsal shaft and angular deformity at the second metatarsophalangeal joint, nor was there an association between medial cortical hypertrophy and metatarsus primus elevatus. A second crossover toe deformity has a peak incidence in women over 50 years of age and has a significant association with both hallux valgus deformities and first metatarsophalangeal degenerative joint disease. Pain is consistently localized to the second metatarsophalangeal joint, and a positive drawer sign is a reliable and consistent physical examination finding. The most reliable radiographic indicator of a second crossover toe is the angle of the second metatarsophalangeal joint in relation to both the hallux and the adjacent third metatarsophalangeal joint angle. There is no relationship between a crossover second toe deformity and the relative length of the second metatarsal, the second metatarsal medial cortex thickness or shaft thickness, metatarsus adductus, radiographic measurements of pes planus, or metatarsus primus elevatus. APPENDIX 1 Technique of intermetatarsal and metatarsophalangeal angular measurements. 11 First, second, and third ray angular measurements are drawn using two mid-diaphyseal reference points on each respective proximal phalanx and metatarsal. These points are placed 1 to 2 cm from the articular surfaces and are connected by a longitudinal line that forms the longitudinal axis of the specific bone. Angle A is the hallux valgus angle, angle B is the 1 2 intermetatarsal angle, angle C is the 2 3 intermetatarsal angle, and angles D and E are the second and third metatarsophalangeal joint angles, respectively.

12 1234 KAZ AND COUGHLIN Foot & Ankle International/Vol. 28, No. 12/December 2007 longitudinal axis of the second metatarsal is drawn using two metaphyseal-diaphyseal reference points. The intersection of these two lines acts as a center of rotation for metatarsal length measurements. Using a compass, two arcs are drawn: one at the distal extent of the articular surface of the first metatarsal and one at distal extent of the articular surface of the second metatarsal. A perpendicular line drawn between the two arcs is measured in millimeters, with a positive value indicating a second metatarsal that is relatively longer than the first. A similar perpendicular line is drawn from the midpoint of the head of the third metatarsal to the arc of the first metatarsal. Appendix 1 APPENDIX 2 Measurement technique of metatarsus primus elevatus. 12 On the preoperative weightbearing lateral radiograph, a line is drawn along the distal dorsal metaphyseal cortex of the first and second metatarsals. A perpendicular line is drawn between these two cortical lines, and the difference between them is measured in mm using an electronic caliper. Normal is considered 8 mm or less. 30 Appendix 3 APPENDIX 4 APPENDIX 3 Appendix 2 Measurement technique of metatarsal protrusion distance. 28 A transverse reference line is made by bisecting two points (one at the most lateral aspect of the calcaneocuboid joint and one at the most medial aspect of the talonavicular joint). The Measurement technique of the metatarsus adductus angle. 14 A line is drawn on the lateral aspect of the foot between the most lateral extent of the calcaneocuboid joint (CC) and the most lateral extent of the fifth metatarsocuboid joint (5MC). A second line is drawn between the most medial extent of the first metatarsocuneiform joint (1MC) and the most medial extent of the talonavicular joint (TN). A mark is made at the midpoint of each of these lines, and these two midpoint marks are connected, thereby bisecting the lesser tarsus. A line perpendicular to this bisecting line is then drawn. The angle that is formed between this perpendicular line and the longitudinal axis of the second metatarsal is the metatarsus

13 Foot & Ankle International/Vol. 28, No. 12/December 2007 CROSSOVER SECOND TOE 1235 (Line C1-C2) as described previously. A second line that bisects the calcaneus (Line D1-D2) is used to form the lateral talocalcaneal angle (Angle 1). Normal is between degrees. 44 APPENDIX 6 Measurement technique of the AP talonavicular coverage angle. 21,33 On the AP weightbearing radiograph, points are marked on the most medial (Point A) and lateral (Point B) margins of the articular surface of the talus, and a line is drawn connecting these two points (Line AB). Similar points are marked on the most medial (Point C) and lateral (Point D) margins of the articular surface on the navicular (line CD), and a line is drawn connecting these two points (Line CD). A perpendicular line is drawn from each of these lines, forming the AP TN coverage angle (Angle x). Normal is 14 +/ 3 degrees. 33 Appendix 4 adductus angle. Normal is 0 15 degrees, mild is degrees, moderate is degrees, and severe is greater than 25 degrees. 14 APPENDIX 5 Measurement technique of the Meary angle and the lateral talocalcaneal angle The Meary angle (lateral talometatarsal angle) is measured according to the technique described by Gould. 22 On the lateral weightbearing radiograph, the longitudinal axis of the talus (Line C1-C2) is established by placing a mark at the halfway point between the superior and inferior surfaces of the talus at the middle of the talus and the neck of the talus, and connecting these two points. A similar method is used to determine the lateral longitudinal axis of the first metatarsal (Line E1-E2), with mid-diaphyseal reference points used to form this axis. The angle formed by these two lines is the Meary angle (Angle 2), with a normal value between 4 and The lateral talocalcaneal angle is measured as described by Shereff. 44 The same talar axis line is used Appendix 5 Appendix 6

14 1236 KAZ AND COUGHLIN Foot & Ankle International/Vol. 28, No. 12/December 2007 APPENDIX 7 Appendix 7 Measurement technique of second metatarsal cortical thickness, shaft thickness, and intramedullary thickness. 24 The length of the second metatarsal is measured from the most distal articular surface to the most proximal articular surface. At the midway point along the long axis of the second metatarsal a transverse line is marked. At this location the medial cortical thickness (MCT), shaft thickness (ST), and intramedullary canal thickness (IMT) are measured with the electronic caliper. REFERENCES 1. Barca, F; Acciaro, AL: Surgical correction of crossover deformity of the second toe: a technique for tenodesis. Foot Ankle Int. 25: , Bhatia, D; Myerson, MS; Curtis, MJ; Cunningham, BW: Anatomical restraints to dislocation of the second metatarsophalangeal joint and assessment of a repair technique. J. Bone Joint Surg. 76- A: , Branch, HE: Pathologic dislocation of the second toe. J. Bone Joint Surg. 19: , Brunet, JA; Tubin, S: Traumatic dislocations of the lesser toes. Foot Ankle Int. 18: , Coughlin, MJ: Crossover second toe deformity. Foot Ankle 8:29 39, Coughlin, MJ: Lesser toe deformities. Orthopaedics 10(1):63 75, Coughlin, MJ: When to suspect crossover second toe deformity. J. Musculoskel. Med. 4:39 48, Coughlin, MJ: Subluxation and dislocation of the second metatarsophalangeal joint. Orthop. Clin. North Am. 20: , Coughlin, MJ: Second metatarsophalangeal joint instability in the athlete. Foot Ankle 14: , Coughlin, MJ: Lesser toe abnormalities. Instr. Course Lect. 52: , Coughlin, MJ; Saltzman, CL; Nunley, JA 2nd: Angular measurements in the evaluation of hallux valgus deformities: a report of the ad hoc committee of the American Orthopaedic Foot and Ankle Society on angular measurements. Foot Ankle Int. 23:68 74, Coughlin, M; Shurnas P: Hallux rigidus. Grading, and longterm results of operative treatment. J. Bone Joint Surgery. 85- A: , Coughlin, M; Mann, R: Lesser toe deformities. In: Mann, R.A.; Coughlin, M.J. (eds). Surgery of the Foot and Ankle. 6th edition, St. Louis, Mosby p.148, Coughlin, MJ: Juvenile hallux valgus. In: Coughlin, M.J.; Mann R.A (eds). Surgery of the Foot and Ankle. 7th edition, St. Louis, Mosby pp , Coughlin, MJ; Mann, RA: Lesser toe deformities. In: Coughlin, M.J.; Mann R.A. (eds). Surgery of the Foot and Ankle. 8th edition, vol 1. St. Louis, Mosby pp , Coughlin, M; Mann, R (eds.): Hallux valgus in surgery of the foot and ankle, 8 th edition. Mosby-Elsevier, Philadelphia, pp , Deland, JT; Sobel, M; Arnoczky, S; Thompson, F: Collateral ligament reconstruction of the unstable metatarsophalangeal joint: an in vitro study. Foot Ankle 13: , Deland, JT; Lee, KT; Sobel, M; DiCarlo, EF: Anatomy of the plantar plate and its attachments in the lesser metatarsal phalangeal joint. Foot Ankle Int.16: , Deland, J; Sung Il-Hoon: The medial crossover toe: a cadaveric dissection. Foot Ankle Int. 21: , Gazdag A; Cracchiolo, A: A surgical treatment of patients with painful instability of the second metatarsophalangeal joint. Foot Ankle Int. 19: , Gould, N: Evaluation of hyperpronation and pes planus in adults. Clin. Orthop. 181:37 45, Gould, N: Graphing the adult foot and ankle. Foot Ankle 2: , Graziano, T: Correction of crossover second toe deformity. Clin. Podiatr. Med. Surg. 13: , Grebing, BR; Coughlin, MJ: Evaluation of Morton s theory second metatarsal hypertrophy. J. Bone Joint Surg. 86-A: , Gross, M; Evanski, PM; Waugh, T: Metatarsalgia from subluxation of the second toe. Foot Ankle 4: , Hak, DJ, Gautsch, TL: A review of radiographic lines and angles used in orthopedics. Am. J. Orthop. 24: , Hansen, ST (eds.): In: Functional Reconstruction of the Foot and Ankle. Lippincott, Philadelphia, pp , , Hardy, R; Clapham, J: Observations on hallux valgus. J. Bone Joint Surg. 33-B: , Hatch D; Burns M: An anomalous tendon associated with crossover second toe deformity. J. Am. Podiatr. Med. Assoc. 84: , Horton, G; Park, Y; Myerson, M: Role of metatarsus primus elevatus in the pathogenesis of hallux rigidus. Foot Ankle Int. 20: , Jahss, MH: Miscellaneous soft tissue lesions. In: Disorders of the Foot. Vol 1. W.B. Saunders, Philadelphia, pp. 843, Johnston, RB; Smith, J; Daniels, T: The plantar plate of the lesser toes: an anatomic study in human cadavers. Foot Ankle Int. 15: , King, DM; Toolan, BC: Associated deformities and hypermobility in hallux valgus: An investigation with weightbearing radiographs. Foot Ankle Int. 25: , Klaue, K; Hansen, ST; Masquelet, AC: Clinical, quantitative assessment of first tarsometatarsal mobility in the sagittal plane and its relationship to hallux valgus deformity. Foot Ankle Int. 15:9 13, Mann, RA; Mizel, MS: Monoarticular nontraumatic synovitis of the MTP joint: a new diagnosis. Foot Ankle 6:18 21, 1985.