lable at ScienceDirect

The Journal of Foot & Ankle Surgery 50 (2011) 272–275

Contents lists avai

The Journal of Foot & Ankle Surgery

journal homepage: www.j fas .org

Original Research

Results of Modified Lapidus Arthrodesis Procedure Using Medial Eminenceas an Interpositional Autograft

Lee Fleming, DPM1, Thomas J. Savage, DPM, FACFAS 2, Matthew H. Paden, DPM, FACFAS 3,Paul A. Stone, DPM, FACFAS 4

1 Submitted during Third Year of Residency, Presbyterian/St. Luke’s Medical Center, Denver, CO2 Private Practice, Aurora; Attending Faculty, Highland's/Presbyterian/St. Luke's Medical Center, Denver, CO3Director of Residency Education, Highland's/Presbyterian/St. Luke's Medical Center, Denver, CO4Director of Research, Highland's/Presbyterian/St. Luke's Medical Center, Denver, CO

a r t i c l e i n f o

Level of Clinical Evidence: 4Keywords:bone graftbunionectomyhallux valgusmetatarsalsurgery

Financial Disclosure: None reported.Conflict of Interest: None reported.Address correspondence to: Thomas J. Savage, DPM

land's/Presbyterian/St. Luke’s Medical Center, 1719 EastE-mail address: metrofoot@aol.com (T.J. Savage).

1067-2516/$ - see front matter � 2011 by the Americdoi:10.1053/j.jfas.2011.02.012

a b s t r a c t

The Lapidus procedure has received wide acceptance as a valuable operation for correcting moderate to severehallux valgus, especially in the presence of hypermobility. However, shortening of the first ray inherentlyoccurs as the first metatarsocuneiform joint cartilage and subchondral bone are resected in preparation forarthrodesis. The purpose of this study was to radiographically compare the degree of shortening of the first raywith and without the use of the first metatarsal medial eminence as an interpositional autograft at the site ofmetatarsocuneiform fusion. Preoperative and postoperative radiographs were measured in 35 consecutivepatients who underwent 37 modified Lapidus procedures for hallux valgus repair. In group A, 20 surgerieswere performed without use of the interpositional autograft, and served as the control. In group B, 14 surgerieswere performed using the medial eminence as an interpositional autograft. The mean amount of first rayshortening was 5.3 � 1.66 mm in group A and 2.69 � 1.56 mm in group B, and this difference was statisticallysignificant (P < .001). All patients progressed to complete union, and the median follow-up was 6 months(range, 4–60). Based on these results, the use of the medial eminence as an interpositional autograft inconjunction with Lapidus arthrodesis resulted in a 49.2% reduction in the amount of shortening of the first rayand proved to be a useful source of readily available bone graft.

� 2011 by the American College of Foot and Ankle Surgeons. All rights reserved.

In 1934, Lapidus described first metatarsocuneiform (MC)arthrodesis for correction of hallux valgus (1). Resection of the met-atarsocuneiform joint (MCJ) allowed for dramatic correction in thesagittal, frontal, and transverse planes. Inherently, however, resectionof the MCJ led to shortening of the first ray, often with resultantoverload of the second ray, development of intractable plantar kera-tosis, lesser metatarsal stress fracture, and predislocation syndrome ofthe second metatarsophalangeal joint (MTPJ), all of which have cometo be known as potential complications of the Lapidus procedure.

In an effort to counteract these complications, Butson (2)described use of the medial eminence from the first metatarsalhead exostectomy as a bone autograft to be used during MC fusion,the graft being positioned between the first metatarsal base andmedial cuneiform and stabilized with Kirschner wire fixation. Butsonreported good to excellent results in 110 (92.44%) of 119 feet with no

, FACFAS, Attending Faculty, High-19th Avenue, Denver, CO 80111.

an College of Foot and Ankle Surgeon

cases of non-union; however, no mentionwas made as to the amountof first ray shortening when the autograft was used (2).

The aim of the investigation that we describe in this report was toquantify the amount of shortening of the first ray after Lapidusarthrodesis with and without use of the first metatarsal head medialeminence as an interpositional autograft at the site of the first MCarthrodesis. We hypothesized that the use of autologous medialeminence with the Lapidus procedure would result in significantlyless shortening and would be adequate without compromising heal-ing of the fusion site.

Patients and Methods

The authors undertook a review of themedical records of consecutive patients whounderwent Lapidus arthrodesis over 87 months, from January 2001 to March 2008. Tobe included in the cohort, a patient had to have undergone a Lapidus arthrodesis, withor without use of the resected first metatarsal medial eminence as an interpositionalautograft at the site of the MC fusion, during the observation period. Consecutivepatients who had undergone Lapidus arthrodesis were identified from the records ofthe senior author (T.S.), using the Current Procedure Terminology code 28297 as thesearch term. The authors procured the identified records and abstracted informationfrom the charts. Exclusion criteria included patients who necessitated any additional

s. All rights reserved.

L. Fleming et al. / The Journal of Foot & Ankle Surgery 50 (2011) 272–275 273

bony procedure on the first or second metatarsal that would effect the measurement ofthe metatarsal protrusion distance. In addition, 3 of 49 radiographs were damagedduring prolonged storage. Incomplete drying prior to filing or moisture during storagecaused films to stick together, and they were damaged while trying to separate them.Thus, we were unable to obtain accurate pre and/or post surgical measurements fromthese few damaged films. These patients alsowere excluded from our study data. Unionof the arthrodesis was defined as radiographic consolidation across the fusion site withan asymptomatic foot, as determined by the authors. First and second metatarsalprotrusion distances were measured on the immediate preoperative and late-termpostoperative weight-bearing anteroposterior radiographs. With regard to the assess-ment of the shortening of the first ray, an arbitrary point of reference was chosen at thedistal navicular cartilage where the medial and intermediate cuneiform joints coincide(Figure 1). A line was thenmeasured from this point to themost distal aspect of the firstand second metatarsal heads and recorded in millimeters. The difference between thelengths of the first metatarsal and second metatarsal was compared preoperatively andpostoperatively (Figures 2 and 3).

Fig. 2. Preoperative measurements of relative lengths of the first and second metatarsals.

Surgical Intervention

After standard prepping, draping, and application of an ankle tourniquet, a singlelongitudinal incision was created medial and parallel to the extensor hallucis longustendon from the medial cuneiform to the midshaft of the hallux proximal phalanx. Theincision was carried through the subcutaneous layer with care to cauterize and/orretract all vascular structures and to retract any nerves. Next, the capsular and peri-osteal incision was performed to expose the first MTPJ and first MCJ. A sagittal saw wasthen used to remove the cartilage and ligamentous attachments from the medial aspectof the medial eminence of the first metatarsal head. The thickness of the removed bonewas typically 1 to 2 mm. Next, a parallel cut with the saw was made to remove theremaining medial eminence with care taken to preserve the sagittal groove plantarlyfor articulation with the base of the proximal phalanx and the tibial sesamoid. Afterremoving any remaining cartilage and soft tissue from the autograft with a rongeur, thebone was then placed on the back table in saline solution to await placement into thefirst MC fusion site, in those cases wherein the interpositional autograft was to be used;otherwise, the eminence was discarded. Next, the lateral release and first MCJ prepa-rationwere performed in a standard fashion. In every case, the sagittal sawwas used forcartilage and subchondral plate removal, keeping the amount of bone resection toa minimum. The opposing surfaces of the first metatarsal base and medial cuneiformwere then fenestrated, and, in those cases wherein the interpositional autograft wasused, the graft bone was positioned at the site of the MC fusion and checked withintraoperative fluoroscopy. Using standard principles of osteosynthesis, the MC fusionwas then stabilized with two 4-mm diameter, partially threaded, cannulated cancellousscrews inserted in a lag fashion from distal to proximal. The wound was then closed inlayers and dressed in standard fashion. The postoperative recovery entailed 6 to 8weeks of non-weight bearing in a below-the-knee cast with crutches to assistambulation.

Fig. 1. Ridge on distal surface of navicular at the intersection of the medial and inter-mediate cuneiforms used as a point of reference to measure metatarsal lengths (circledarea).

Results

Thirty-nine consecutive patients were identified as eligible forinclusion in the retrospective cohort. One surgeon (senior author, T.S.)performed all of the operations. One patient, who underwenta concomitant Reverdin osteotomyof thefirstmetatarsal at the time ofthe Lapidus procedure, was excluded based on the aforementionedcriterion. Three patients with missing or damaged radiographs werealso excluded. Therefore, 46 procedures (35 patients) were included inthe analyses. Nine patients underwent a second (bilateral) procedureat a separate surgical encounter. A comparison of the demographic andoutcomevariables, based onwhether the interpositional autograftwasused, is depicted in Table 1. Overall, 46 Lapidus procedures were per-formed in 37 (80.4%) women and 9 (19.6%) men. Overall, themean ageof the cohort was 42.4 years (range, 13–74 years). A total of 20 (43.5%)procedures were performed without use of the interpositional medialeminence autograft, whereas 26 (56.5%) procedures were undertakenwith use of the autograft. Group A (MC fusion without autograft) had20 Lapidus procedures performedon13 (65%)womenand7 (35%)menwith amean age of 40.4 years (range,14–74 years). Group B (MC fusionwith autograft) had 26 Lapidus procedures performed on 25 (96.2%)women and 1 (3.8%) manwith a mean age of 43.9 years (range, 13–72years). Themean shorteningwas 5.3�1.66mm for groupA (MC fusionwithout autograft), and 2.69 � 1.56 mm for group B (MC fusion withautograft), and this difference was statistically significant (P < .001).There was 1 delayed union (6 months), 1 malunion (Hallux varus

Fig. 3. Postoperative measurements of relative lengths of first and second metatarsals.Note the presence of the medial eminence at the fusion site.

L. Fleming et al. / The Journal of Foot & Ankle Surgery 50 (2011) 272–275274

developed), and no nonunions associated with these procedures. Weobserved 4 cases of postoperative wound cellulitis, 2 in each group,each of which resolved with use of oral antibiotics without otherintervention. In this study, the modified Lapidus procedure using themedial eminence of the first metatarsal as autograft resulted in anaverage reduction of shortening of 2.69 mm. This corresponds toa 49.2% overall reduction of shortening compared with the controlgroup. The median duration of follow-up was 6 months (range, 4–60months), with no evidence of the development of lessermetatarsalgia.

Discussion

We observed less first ray shortening when the resected medialeminence of the first metatarsal head was used as an

Table 1Comparison of clinical variables by intervention (N ¼ 46 procedures)*

Variable Lapidus withoutAutograft (n ¼ 20)

Lapidus withAutograft (n ¼ 26)

Age (years) 40.4 (14–74) 43.9 (13–72)SexMale 7 (35.0%) 1 (3.8%)Female 13 (65%) 25 (96.2%)

Bone healing complication 0 2 (4.3%)Shortening (mm) 5.3 � 1.66 2.69 � 1.56Duration of follow-up (months) 8 � (4–60) 5 � (4–30)

* Results shown as mean � standard deviation of median and range for continuousdata, and count (%) for categorical data.

interpositional autograft at the first MC arthrodesis site. Whenconsidering that one of the main complications of first MC fusion isexcessive shortening of the first ray that could result in lessermetatarsalgia, the findings of this report could be useful tosurgeons performing Lapidus arthrodesis. This finding is significant.Our review of the literature showed measurements of first rayshortening after the Lapidus procedure similar to those observed inour control group. In fact, we observed 5.3 � 1.66 mm of shorteningwhen the autograft was not used, whereas Saffo et al (3) noted 6mm, Sangeorzan and Hansen (4) noted 5 mm, Catanzariti et al (5)noted 4.7 mm, and McInnes and Bouch�e (6) noted 4.5 mm ofshortening after Lapidus arthrodesis.

Anecdotally, the medial eminence of the first metatarsal head hasbeen considered suboptimal bone for use as autograft. However, theresults of this study showed no clinically or statistically significantdifferences in the time to healing, or bone healing complications,between the test (autograft) group and the control (no autograft)group. Some advantages of using the medial eminence as graftmaterial over allogeneic or other autograft bone are that the medialeminence is readily available through the same incision, and there isno additional cost associated with using this bone graft. One disad-vantage is that the size of the autograft is limited to the amount ofbone removed, after any cartilage is removed from the fragment. Inour experience, the size of the autograft ranges from 2 to 5 mm inthickness. Understanding that much of the graft thickness is lostduring the healing process, it is important to fashion the autograft sothat it is as thick as it can possibly be, without over aggressivelystaking the metatarsal head (which could predispose to halluxvarus).

In this investigation, we determined the difference between thelength of the first and second rays by comparing metatarsal protru-sion. This measurement involved longitudinally bisecting the first andsecond metatarsals and using the point where the 2 bisections met asthe center of an arc. Two arcs were drawn, one arc at the most distalaspect of the first metatarsal head and the other at the most distalaspect of the second metatarsal head. The length difference betweenthese 2 arcs was measured and designated the metatarsal protrusiondistance. Because an ideally corrected first intermetatarsal angleapproaches 0�, and the bisections of the first and second metatarsalswould never intersect if the long axis of each metatarsal was parallelto the other, thismeasurement could not be used in our study. Instead,we arbitrarily identified the most distal aspect of the first and secondmetatarsal heads and calculated the difference between the 2distances.

Like many retrospective clinical investigations, we realize thata number of methodological shortcomings could have influenced ourfindings. It is important to note that although we accounted forpostoperative complications that we were able to identify in themedical records, we did not measure a subjective outcome, such aspatient satisfaction with the results of the surgery. Moreover, thesame surgeons who performed the operations abstracted the datafrom the records and measured the radiographs, which could haveimparted some bias. Similarly, measurement bias could have influ-enced the arbitrary designation of the most distal margin of the firstand second rays. Finally, the decision to use the autograft was madesolely at the discretion of the surgeon, and treatment was not allo-cated in a random fashion.

In conclusion, we have described a modification of the Lapidusprocedure for hallux valgus using the medial eminence as autograft tominimize shortening of the first ray. We have shown the medialeminence of the first metatarsal to be viable bone, readily available,able to withstand compressive forces, and able to provide statisticallysignificant reduction of shortening of the first ray. We believe that theresults of this investigation could be useful to surgeons interested in

L. Fleming et al. / The Journal of Foot & Ankle Surgery 50 (2011) 272–275 275

designing and undertaking future prospective studies and random-ized controlled trials that focus structural changes associatedwith useof the Lapidus procedure.

References

1. Lapidus PW. Operative correction of the metatarsus varus primus in hallux valgus.Surg Gynecol Obstet 54:183–191, 1934.

2. Butson ARC. A modification of the Lapidus operation for hallux valgus. J Bone JointSurg (Br) 62-B(3)350–352, 1980.

3. Saffo G, Wooster MF, Stevens M. First metatarsocuneiform joint arthrodesis: a five-year retrospective analysis. J Foot Surg 28:459–465, 1989.

4. Sangeorzan BJ, Hansen ST. Modified Lapidus procedure for hallux valgus. Foot Ankle9(6):262–266, 1989.

5. Catanzariti AR, Mendicino RW, Lee MS, Gallina MR. The modified Lapidusarthrodesis: a retrospective analysis. J Foot Ankle Surg 38(5):322–332, 1999.

6. McInnes BD, Bouch�e RT. Critical evaluation of the modified Lapidus procedure.J Foot Ankle Surg 40(2):71–90, 2001.