Freiberg Disease 

Updated: May 20, 2019
Author: Shayne D Fehr, MD, FAAP; Chief Editor: Vinod K Panchbhavi, MD, FACS 



In 1914, Alfred H Freiberg first described the painful collapse of the articular surface of the second metatarsal head.[1] He reported six cases of young women presenting with a painful limp and discomfort localized to the second metatarsal. All six patients had similar radiographic findings, which showed collapse of the articular surface of the second metatarsal head. In three patients, intra-articular loose bodies were seen. Of the six women, four were younger than 18 years.

Because Freiberg believed that the etiology most likely was trauma, he used the term infraction.[1] He postulated that the condition was partially caused by excessive length of the second metatarsal. which, when combined with an ineffective first ray complex, led to an overload of the second metatarsal and subsequent articular collapse.

Since Freiberg's original description, multiple authors have written about this uncommon condition, which has since come to be known as Freiberg disease (first described as Freiberg infraction but also referred to, inaccurately, as Freiberg infarction).

Although Freiberg disease was originally described more than a century ago, the etiology and the most appropriate treatment remain controversial. A trial of conservative treatment can be implemented for most patients presenting with Freiberg disease. If such treatment fails, multiple surgical options exist.


In its juvenile osteochondrosis form, Freiberg disease is a degenerative process involving the epiphyses resulting in osteonecrosis of subchondral cancellous bone. If the process is altered in such a way as to restore normal physiology, this may be followed by regeneration or recalcification. If not, the process continues to subchondral collapse and eventual fragmentation of the joint surface.[2]


In his original description, Freiberg favored a traumatic etiology; however, he admitted that trauma itself could not reliably explain all cases.[3] Despite several thoughtful investigations into the etiology of Freiberg disease, no consensus exists as to what type of disease process this peculiar affliction represents. This lack of consensus is reflected in the numerous terms used in the literature to describe this process.[4]

Freiberg disease in adolescents is thought to belong to a group of related diseases involving growth disturbances of the epiphysis or apophysis, collectively termed the osteochondroses. Although it is considered to represent an interruption of normal growth processes, the specific events or factors that incite the cascade leading to articular collapse are unclear. Among the more common articular osteochondroses are Köhler disease of the tarsal navicular, Legg-Calve-Perthes disease of the femoral head, and Panner disease of the humeral capitellum.[5]

Although regarding Freiberg disease as an osteochondrosis makes sense, it does not fully explain the adult-onset form of the disease, which may reflect a different process altogether, albeit radiographically similar one. Some authors consider the cause to be multifactorial. Unlike other osteochondroses, Freiberg disease is not commonly associated with steroid use, alcoholism, or smoking. Most current theories center on whether the initial insult is primarily traumatic or vascular. Infection has essentially been eliminated as a significant factor.

Vascular insult

Radiographic changes consistent with avascular necrosis have led some authors to suggest that the inciting event is an injury to the vascular supply to the metatarsal head. Whether this is the result of a direct vascular injury or of repetitive injury to an area that has an inherently poor blood supply is unknown. Disorders that may affect blood supply (eg, diabetes mellitus, systemic lupus erythematosus, and hypercoagulability) have been suggested factors in the development of Freiberg disease.[6]

With the aim of better understanding the blood supply to the metatarsal heads, several cadaveric investigations have been performed. Huber described the variability of the dorsal arteries in 200 feet, finding that 65% of those feet exhibited limited collateral anastomosis.[7] Leung and Wong described as many as seven different patterns of supply originating from the first webspace dorsal artery.[8]

Wiley and Thurston noted variation, finding that 33% of specimens studied lacked a second metatarsal artery, with collaterals from the first and third metatarsal arteries supplying the second metatarsal head.[9] Viladot and Viladot noted that the metatarsal head receives its vascular supply from small penetrating vessels at the joint capsule.[10]

These studies demonstrated that the vascular supply to the metatarsal heads can be quite variable. Overall, there appears to be a trend in which the second and third metatarsals receive a less consistent blood supply than the other metatarsals do. These studies may suggest that some patients may have a greater risk for the development of Freiberg disease based on their anatomic variances.

Others have cited reports of iatrogenic avascular necrosis of the second and third metatarsal heads following elective forefoot surgery as indirect evidence that a disturbed blood supply may be at least partially responsible for the development of Freiberg disease.[11]

Traumatic insult

Some authors, however, favor trauma as the predominant etiologic agent in the development of Freiberg disease, whether in the form of a single acute injury or in the form of multiple repetitive microinjuries. Various authors have suggested that altered kinematics around the forefoot may predispose some patients to injury as a result of abnormal loading and may ultimately contribute to the development of Freiberg disease.[4, 12] Several different local mechanical factors have been implicated.

One anatomic variant often implicated in Freiberg disease is a long second metatarsal. In his original description, Freiberg postulated that a long second metatarsal in combination with altered first ray mechanics eventually leads to overload of the second metatarsophalangeal (MTP) joint. He theorized that with forcible impact in feet predisposed by weakness of the toe flexors (especially the first toe), the longer second toe would be susceptible to injury. However, he did note that only four of his six original cases had a discrete history of trauma.[1]

Similarly, in their evaluation of various etiologic factors implicated in Freiberg disease, Stanley et al found that as measured from standing radiographs, the affected ray was the longest in 85% of the feet.[4] The authors believed this finding to be important. They noted that only five of 33 feet (15%) in their study had a discrete history of trauma.

Of the metatarsals, the second and third are the least mobile. This has led some investigators to conclude that these metatarsals, because of their relative inflexibility and increased load transmission, are at increased risk for sustaining repetitive microtrauma. Similarly, Smillie considered Freiberg disease to be a repetitive stress injury, analogous to a march, or stress, fracture.[13] He believed that concentration of stress in the trabecular bone at the dorsal aspect of the metatarsal head eventually leads to collapse.

In their summary of various etiologic factors, Stanley et al evaluated pressure under the metatarsals in patients diagnosed with Freiberg disease, finding that only five of 33 feet showed abnormally high pressures at the affected site.[4] However, in 16 patients (48%), the greatest measured pressure was indeed at the site where Freiberg disease was present.

McMaster proposed a possible mechanism of injury for the development of Freiberg disease.[14] On the basis of similarities between lesions observed in hallux rigidus and Freiberg disease, he suggested that the typical location of the lesions could be explained on the basis of mechanical impingement between the base of the proximal phalanx and the dorsum of the metatarsal head in forced dorsiflexion. He postulated that the relative inflexibility of the second and third metatarsals contributed to the development of lesions at these particular sites.

In an attempt to test this hypothesis, Helal and Gibb induced joint incongruity by creating artificial effusions.[15] Viscous silicone was injected into the MTP joints of cadaveric feet. Joint congruity was then radiographically assessed in varying degrees of dorsiflexion and plantar flexion.

Helal and Gibb observed dorsal impingement of the metatarsal head and the proximal phalanx under these conditions.[15] They attributed the dorsal impingement to joint incongruity induced by the effusion. They believed that the impingement occurred dorsally because the MTP joint capsule is inherently thicker dorsally than it is ventrally. However, whether or not an effusion plays a significant role in the development of Freiberg disease in vivo is unclear.

In an attempt to evaluate trauma in the development of Freiberg disease, Braddock subjected intact cadaveric second MTP joints to axial loads. In two of the 10 specimens, he was able to create lesions closely resembling those seen in Freiberg disease. The stage of epiphyseal maturation in the two specimens resembled that commonly seen in adolescent-onset Freiberg disease. This led him to conclude that trauma to the second metatarsal at the proper phase of epiphyseal maturation could produce lesions similar to those of Freiberg disease.[16]

Others have considered the disorder's predilection for women to be indirect evidence that trauma plays an etiologic role. Theories are based on male-female differences in the selection of shoe wear; presumably, high-heeled shoes subject the metatarsal heads to repetitive injury and increased pressure transmission. Hoskinson reported the development of Freiberg disease in three females after initial attempts at wearing high-heeled shoes.[17] Anecdotal experience aside, there is no direct evidence linking the disease with the wearing of high heels.

Nguyen et al observed seven cases of Freiberg disease occurring in older patients with diabetes.[18] They questioned whether trauma to the metatarsal heads as an indirect result of a peripheral neuropathy could result in the development of Freiberg disease. They postulated that intrinsic motor weakness, such as is often seen with peripheral neuropathy, can lead to extension of the toes at the MTP joint, resulting in an increase in weightbearing by the metatarsal heads, repetitive injury, and subsequent collapse.

Of the seven patients studied by Nguyen et al, three had a documented neuropathy.[18] The authors did not comment on whether the patients had protective sensation or if changes of Charcot arthropathy were present in the other joints of the foot.

In summary, despite more than a century of description, the exact etiology of Freiberg disease is still unknown. Thus, it is most likely multifactorial, including vascular and traumatic insults. Certain patients may be anatomically predisposed by mechanical, vascular, and developmental factors. Whether the process is the same for older patients as it is for adolescents is unknown.

The relative infrequency of the disease, as well as the variable presentation regarding age and injury, makes the study of various etiologies challenging. In the future, genetic or other variables that play a role in the development of Freiberg disease may be identified. Further investigation is needed.


The true incidence of Freiberg disease has not been established. Some cases are asymptomatic, and others may resolve spontaneously before treatment is sought. Freiberg disease appears to be an uncommon condition, as evidenced by the small number of patients in most series that are reported in the literature.

Consistent with Freiberg's original description, the disease most commonly is seen in young women. In a review of 275 reported cases, Katcherian found an overall male-to-female ratio of approximately 1:5.This female preponderance is unusual among the osteochondroses as a whole, in which males are more affected.[19]

In more than 95% of cases, the lesion is found in the second or third metatarsal, with the second metatarsal being affected more often. However, any of the metatarsals may be involved, though the first and fifth are rarely affected.[20] Almost always, there is a single lesion in one foot. Bilateral involvement has been reported to occur in fewer than 10% of patients.[21]

The individuals most commonly affected range in age from adolescence through the second decade of life. However, Freiberg disease can occur at any age, with an age range of 8-77 years reported in the literature.




Patients who have Freiberg disease typically present with complaints of activity-related forefoot pain. Walking alone is often sufficient to cause pain. Some patients describe an extended (months to years) history of chronic forefoot pain with episodic exacerbation, whereas others present with pain of recent onset that is related to a specific injury or event. A history of trauma may not be noted. Patients may present with stiffness and a limp. The pain is often vague and poorly localized to the forefoot. Some patients describe the sensation of a small, hard object under the foot.

Physical Examination

Physical examination typically reveals a limited range of motion (ROM), swelling, and tenderness with direct palpation of the metatarsophalangeal (MTP) joint. In early stages of the disease, MTP tenderness may be the only finding. In later stages, crepitus or deformity may be present. A skin callus may be seen on the plantar surface of the affected metatarsal head.



Diagnostic Considerations

The most common differential diagnoses include metatarsalgia, Morton neuroma, stress fracture of the metatarsal, and synovitis.

Differential Diagnoses



Laboratory Studies

On the basis of the clinical presentation and imaging findings, laboratory studies and other investigations may be indicated to rule out other etiologies of pain or deformity. In particular, infectious, rheumatologic, and oncologic processes may cause pain and abnormal imaging of the metatarsal head.

Imaging Studies


Depending on the stage of the disease, radiographs may show only sclerosis and widening of the joint space (early), with complete collapse of the metatarsal head and fragmentation later. Osteochondral loose bodies may be seen late in the disease as well. Oblique views may be especially useful for achieving a full appreciation of subtle changes early in the disease. One study advocated the use of radiographs to assess musculoskeletal foot conditions in women related to poorly fitting shoes.[22]

Occasionally, patients are completely asymptomatic, with changes noted on radiographs taken for other reasons. Whether these patients later develop symptomatic Freiberg disease is not known.

Bone scintigraphy, MRI, and CT

The use of bone scanning has been described with photopenia in the early stages of the disease, with intense uptake later as the head is reconstituted or revascularized. Although bone scintigraphy has been used in the study of Freiberg disease, its value as a diagnostic or prognostic tool is unknown.[21]

Magnetic resonance imaging (MRI) is helpful in detecting early Freiberg disease not visualized on plain radiographs.[23] MRI may demonstrate hypointense signal in the epiphysis on T1 images and mixed hypointense and hyperintense signals on T2 images. Flattening of the metatarsal head may be identified as well.[6]

MRI has been advocated by some physicians as helpful for preoperative evaluation, especially if an osteotomy is planned. One study demonstrated that three-dimensional (3D) computed tomography (CT) was useful in characterizing the extent of osteonecrosis in a lesion.[24]

Histologic Findings

Histologic examination of tissue from resected specimens has produced varied findings. Different studies have found bone resorption and new bone formation, depending on the stage of the disease.[25]

Young et al described separation of the deeper layers of the hyaline cartilage in a 55-year-old man with Freiberg disease.[26] The separation had occurred in close proximity to the zone of mineralization, and avascularity was not evident. The authors concluded that in this one case, a traumatic shear or compression-type injury was likely to have been responsible, as opposed to some type of vascular insult (avascular necrosis).


Several staging schemes have been described. Most are based on radiographic appearance, including the amount of collapse and the presence or absence of secondary degenerative changes. As is frequently the case with other musculoskeletal conditions, multiple classification systems exist, with varying levels of validation.[24] Nevertheless, the classification scheme developed by Smillie in 1967[27] is the most often quoted system and divides the radiographic changes into the following five stages (see the images below):

  • Stage I - The earliest sign is fissuring of the epiphysis; radiographic changes at this stage may be so subtle that they are missed with routine radiographs
  • Stage II - Later central depression of the articular surface becomes evident as subchondral cancellous bone is resorbed; the articular cartilage hinges on an intact plantar bridge
  • Stage III - The central depression is seen to be resulting in medial and lateral projections at the margins; the plantar hinge remains intact at its plantar isthmus
  • Stage IV - This stage demonstrates that the central portion has sunk below the surface and is free of the plantar hinge, thus becoming a loose body; fractures of the medial and lateral projections are present, with folding of the projections over the central loose body
  • Stage V - This final stage shows marked flattening and deformity of the metatarsal head with secondary degenerative changes; the central loose body may have been resorbed at this stage; the shaft of the metatarsal becomes thickened and dense
Early stage I-II lesion of Freiberg disease, best Early stage I-II lesion of Freiberg disease, best seen on oblique radiograph.
Stage III Freiberg disease with advanced flattenin Stage III Freiberg disease with advanced flattening.
Stage IV Freiberg disease with articular collapse Stage IV Freiberg disease with articular collapse and loose body formation.
Stage V Freiberg disease with advanced degenerativ Stage V Freiberg disease with advanced degenerative changes involving metatarsal head and proximal phalanx.


Approach Considerations

There remains controversy as to what represents the most appropriate approach to treatment of Freiberg disease. Current recommendations have been based on small series of patients treated by various methods, and there is no clear therapeutic consensus. 

A trial of conservative treatment can be implemented for most patients presenting with Freiberg disease. Whereas some stage I, stage II, and stage III lesions (see Staging) may resolve spontaneously, patients who do not respond to conservative measures may require surgery, as may patients with stage IV and stage V lesions. Aside from the recommendation of conservative management in the pediatric population and in patients with early-stage lesions, no true contraindications to treatment of Freiberg disease exist.

Further investigation through prospective, multicenter analysis would best guide future treatment options.

Medical Therapy

Various suggestions regarding conservative therapy for Freiberg disease have been put forth in the literature, depending on the stage and the acuteness of the onset of pain. A period of appropriate nonoperative conservative management is indicated for all patients presenting with Freiberg disease. Regardless of the treatment method chosen, the goal of therapy is to rest the joint to allow inflammation and mechanical irritation to resolve.[28]

In patients presenting with severe pain of an acute nature, a non-weightbearing cast may provide sufficient relief during the acute phase. In other persons, a short leg walking cast or hard-sole shoe may be more appropriate. Typically, the initial period of immobilization lasts 4-6 weeks.

In patients with chronic complaints, less restrictive options, such as shoe modifications in the form of inserts with metatarsal bars or pads, rigid shanks, or a rocker bottom, may be helpful. Activity modification during exacerbations may help to prevent the aggravating symptoms of pain and swelling.[6]

Medications may be helpful for pain but should not be used to allow increased activity. Nonsteroidal anti-inflammatory drugs (NSAIDs) are used commonly for pain, but cautious use is recommended. Although sufficient human studies are lacking, NSAIDs have been shown to delay stress fracture healing in rats.[29]

Hoskinson reported on the long-term results (average, 12 years) of conservative treatment on a series of 16 patients.[17] Eleven of the 16 patients were completely pain free at final follow-up; however, all 16 had some restriction in range of motion (ROM).

Surgical Therapy

The progression of Freiberg disease is variable with regard to time course and severity. Whereas some stage I, stage II, and stage III lesions may resolve spontaneously, patients who do not respond to conservative measures and patients with stage IV and stage V lesions may require surgery. Smillie believed that it was possible to restore joint congruity in early lesions (stages I-III) with surgery.

Several surgical options have been advocated in the past. Simple debridement and loose body removal were originally described by Freiberg. Other described procedures, including various osteotomies, elevation of the depressed metatarsal head with bone grafting, core decompression, metatarsal head excision, shortening of the metatarsal, proximal phalanx hemiphalangectomy, total small-joint arthroplasty, and various combinations of the above, have been implemented.

No consensus exists as to which surgical procedure is the most appropriate for patients with symptomatic Freiberg disease. Common to all of these procedures is the goal of diminishing pain and restoring joint function. However, most studies have included small numbers of patients and have not been stratified by age or stage of the disease, making it difficult to draw conclusions about the potential effectiveness of these procedures. The relative rarity of the condition makes it difficult to perform prospective analyses.

Helal and Gibb suggested one pattern for management of Freiberg disease. They suggested grafting and elevation of the collapsed articular surface for patients with stage I or stage II disease. For later-stage lesions (stages III-V), they suggested tailoring the treatment to the patient's symptoms. They recommended using an osteotomy to treat patients with pressure metatarsalgia and using replacement arthroplasty to treat patients suffering from arthritic symptoms of pain with joint motion.[15]

Although these recommendations may provide guiding principles, it is important to recognize that at present, there is no consensus as to which procedure works best for all patients. In general, if conservative treatment fails, the least destructive and invasive procedures should be considered first for patients with early-stage disease, with the more invasive joint-altering procedures reserved for advanced cases or for patients in whom other forms of treatment have failed.


Freiberg's original monograph reported that two of his six patients had good results with debridement. The report only infers that these two patients presented with advanced-stage disease on the basis of the presence of loose bodies.[1] Subsequent authors also advocated simple debridement as an effective treatment for Freiberg disease.[30] However, staging of the lesions, as well as detailed assessment of results, has not commonly been reported, making it difficult to make recommendations based on the currently available data.

Erdil et al reported on 14 patients with advanced-stage disease of the second metatarsal; they found joint debridement and metatarsal head remodeling to be safe, simple, and effective in alleviating symptoms.[30]

Simple debridement can be combined with other procedures. Hoskinson described 12 patients treated with various surgical procedures, including excision of the metatarsal head (n=4), hemiphalangectomy (n=4), and debridement with loose-body excision (n=4). He believed that the best results were seen following simple debridement, but he warned about drawing conclusions from such limited numbers.[17]

Bone grafting

In an attempt to restore joint congruity, Smillie described a procedure in which a cancellous bone graft was used to elevate the depressed metatarsal articular surface. The technique involved creating a slot in the metatarsal shaft through which the sclerotic bone could be drilled. The metatarsal articular surface was then elevated and supported by a bone graft. The authors recommended this procedure for stage I, II, or III lesions in which an intact cartilage flap is present.[13, 27]

In 1987, Helal and Gibb reported on 25 patients with Freiberg disease; 11 of these patients were treated with a modification of Smillie's original procedure. The authors reported eight patients to be clinically and radiographically normal at 3-8 years following surgery. Three patients demonstrated expanded metatarsal heads radiographically, and two of the three reported pain with running or wearing high-heeled shoes.[15]


Several different osteotomies have been described. The common goal of all osteotomies is to redirect the loading of the metatarsophalangeal (MTP) joint away from the damaged area of the articular surface. The two basic procedures are dorsal closing wedge osteotomy and shortening osteotomy.[31]

Dorsal closing wedge osteotomy

Gauthier and Elbaz were the first to describe a dorsal closing wedge osteotomy for the treatment of Freiberg disease.[20] They reported results from 53 patients who were treated with a dorsiflexion osteotomy of the neck of the metatarsal. As described, the dorsal closing wedge osteotomy reoriented the intact cartilage on the plantar surface to articulate with the base of the proximal phalanx. The authors reported stable results with no complications. In 35 of the 53 cases, the postoperative arc of motion averaged 80°.

Al-Ashhab et al, in a series of 10 patients, showed that dorsiflexion osteotomy of the metatarsal head was a simple procedure that was capable of good results in stage IV and stage V disease.[32]

Chao et al reported results from 13 patients with all stages of Freiberg disease who were treated with dorsal closing wedge osteotomy combined with synovectomy and debridement.[33] Temporary fixation with Kirschner wires (K-wires) was utilized in all patients. The patients all had follow-up visits at an average of 40 months.

The results, as graded by the American Orthopaedic Foot and Ankle Society (AOFAS) lesser toe metatarsophalangeal-interphalangeal scale, were as follows[33] : Four patients reported excellent results, seven good results, and two poor or fair results. Metatarsal shortening averaged 2.1 mm, with passive ROM decreased by an average of 15° of flexion and 8° of extension. One patient, who underwent 3 mm of shortening, experienced transfer metatarsalgia.

Kinnard and Lirette reported on 15 patients who were treated with dorsiflexion osteotomy, including several patients with advanced-stage disease.[34] They reported complete pain relief, with only three patients experiencing mild discomfort with athletic activities. Loss of extension averaged 10°, and flexion loss averaged 15°. Metatarsal shortening averaged 2.5 mm, with no cases of transfer metatarsalgia. No major complications were reported.

Shortening osteotomy

Another osteotomy reported for treatment of Freiberg disease is the shortening osteotomy. The basis for the use of a shortening osteotomy is the observation that when involved, the second metatarsal often is the longest of the metatarsals. Several authors believe that this subjects the involved metatarsal to repetitive injury and abnormal loading. With shortening, overloading of the metatarsal is reduced, as are symptoms.

Smith et al described a shortening osteotomy in which the metatarsal was shortened approximately 4 mm.[35] Of the 16 patients treated in their series, 15 had complete pain relief; however, seven of the 16 patients experienced stiffness of the involved ray, with four patients unable to place the toe flat when standing. Five patients graded their result as excellent, nine said that they were very pleased with the outcome, one was satisfied with it, and one was unhappy with the result.

Two minor complications occurred, with one sinus requiring reexploration and one hardware failure occurring despite union. Advantages cited by the authors included the ease of the procedure, avoidance of damage to the metatarsal head, and apparent remodeling of the articular surface, as seen radiographically in most cases.[35]


Resection arthroplasty

Resection arthroplasty, though advocated in the past, has fallen out of favor for the initial treatment of Freiberg disease. Open and arthroscopic techniques have been described.[36] Two of the described methods are resection of the base of the proximal phalanx or of the metatarsal head. Resection can be combined with soft-tissue interposition arthroplasty or even syndactylization of the toes, or it can be performed without these other procedures.

Hoskinson reported on eight patients treated with resection arthroplasty, both hemiphalangectomy and resection of the metatarsal head, and found that only three had a satisfactory result, with residual symptoms and deformity limiting the remaining five.[17]

Resection arthroplasty is an inherently destructive procedure. With these procedures, several authors have expressed concerns about the development of transfer lesions as a result of rendering the affected metatarsal incompetent. Additionally, potential complications from any of the resection techniques include progressive hallux valgus and excessive shortening.

Total small-joint arthroplasty

Total small-joint arthroplasty using a silicone prosthesis has also been described for the treatment of Freiberg disease. Potential complications are similar to those for resection arthroplasty; other potential problems, inherent in the implants themselves, include synovitis, infection, and dislocation.

Several implants are available commercially. Most of these implants originally were developed for use in the hand. Potential advantages over resection arthroplasty include maintenance of length, improved joint motion, and better weight distribution if condyles are preserved. Newer MTP-specific implants may prove to be better than previous implants, but long-term data are lacking.[37]

Other options for replacement arthroplasty, including a titanium hemiarthroplasty and total ceramic arthroplasty, have been explored in recent years for high-stage lesions. Similar to silicone implants, these newer devices present possible complications, including implant loosening, bone erosion, infection, and a stiff, floating toe.[38, 39]

Interposition arthroplasty

As a consequence of the lack of consistent results with both resection and small-joint arthroplasty, there has been some interest in interposition arthroplasty. Unfortunately, this procedure too has yielded mixed, unreliable results.[40, 36, 41]

Other procedures

Freiberg and Freiberg described core decompression for the treatment of early-stage lesions.[25] In their procedure, the metatarsal head is drilled multiple times with a .045-in. Kirschner wire (K-wire), with satisfactory results (somewhat analogous to drilling for osteochondritis dissecans).

Maresca et al described arthroscopic drilling in a patient with bilateral stage II disease, with evidence of restoration of the joint surface and satisfactory results at 2 years.[42] Although such results are encouraging, further investigation is warranted into the potential usefulness of such procedures.

Miyamoto et al performed an osteochondral plug transplantation for late-stage Freiberg disease in four female patients (average age, 12 years); the osteochondral plug was harvested from a non-weightbearing site of the upper lateral femoral condyle of the ipsilateral knee.[43]

In this study, the average AOFAS score improved from 70.8 points preoperatively to 97.5 points postoperatively.[43] Magnetic resonance imaging (MRI) at 6 months after surgery showed an osteochondral plug–subchondral bone interface, but healing of the plug was confirmed at 12 months after surgery in all patients. At 12 months after surgery, two patients had a normal International Cartilage Repair Society Cartilage Repair Assessment Score, and two had a nearly normal score.