Medial Femoral Condyle Periosteal Free Flap for the Treatment of Talus Avascular Necrosis

Methods

Surgical Technique

The initial part of the surgical procedure was performed by the orthopedic team. At this point, a standard diagnostic arthroscopy was performed using standard technique to assess for collapse of the talar dome. If present, the revascularization procedure was contraindicated. Intra-articular pathology was also addressed accordingly if present.

A direct anterior approach to the ankle was then performed to expose the deep neurovascular bundle (NVB) and nonarticular surface of the talus. Care was taken to avoid any unnecessary soft tissue stripping, which can further compromise the blood supply to the talus. Once the NVB was identified, a core decompression was performed while attempting to debride any areas of necrotic bone using a cannulated drill technique (Figure 1). A guidewire for a 5.0-mm drill bit was first placed from anterolateral at the nonarticular surface to the posteromedial aspect of the talus. Appropriate placement was confirmed with fluoroscopic guidance. Once confirmed, the guidewire was drilled using appropriate technique to avoid thermal necrosis. This path served as the docking site for the MFC periosteal free flap. Additional core decompressions were performed using a 3.5-mm cannulated drill bit (Figure 2). If any cysts were present, these were curetted and bone grafted accordingly from the medial femoral condyle, which is preferred by the senior author (Figure 3). In those patients who also presented with AVN of the distal tibia, a core decompression with debridement of devitalized bone and subsequent bone grafting was performed through the same incision.

OPEN IN VIEWER

Figure 1.

(A, B) A guidewire was placed from anterolateral at the nonarticular surface to the posteromedial aspect of the talus. (C) It was then overdrilled using the Arbeitsgemeinschaft für Osteosynthesefragen (AO) technique with a 5.0-mm cannulated drill bit to create a core decompression tunnel/docking site for the flap.

OPEN IN VIEWER

Figure 2.

(A, B) Additional guidewire placement was dictated by the location of avascular necrosis as noted on preoperative magnetic resonance imaging. (C) It was then overdrilled using the Arbeitsgemeinschaft für Osteosynthesefragen (AO) technique with a 3.5-mm cannulated drill bit.

OPEN IN VIEWER

Figure 3.

(A,B) If talar cysts were present, additional curettage and bone grafting was performed in a retrograde fashion.

Care was then transferred to the plastic and microvascular surgery team for harvest of the medial femoral condyle graft (Figure 4). A longitudinal incision was made at the posterior aspect of the vastus medialis and carried down to the level of the fascia, which was incised longitudinally, followed by retraction of the vastus medialis anteriorly. The descending genicular artery (DGA) and vein were then encountered and traced proximally to the superficial femoral artery and distally to the medial femoral condyle. A 15-blade scalpel was used to harvest a periosteal graft, which was then mobilized completely on the descending genicular artery and vein pedicle. The vessels were clipped and divided at their origin, which was at the level of the adductor hiatus. If talar cysts were present on preoperative imaging, the orthopedic surgeon harvested cancellous graft from the medial femoral condyle at this time to fill the defects in the talus (n = 2). The tourniquet was then deflated to evaluate for any bleeding, and the donor site was closed.^1,7

OPEN IN VIEWER

Figure 4.

(A) Surgical dissection of the descending geniculate artery to the medial femoral condyle (MFC) periosteum. (B) Harvested MFC periosteal graft. (C) Anastomosis of the periosteal free flap to the anterior tibial artery and docking into the talus.

The periosteum was sutured onto itself using a 5-0 chromic to form a cylinder with the inner “cambium” layer now facing out. The vascularized graft was then transferred into the 5.0-mm core decompression site in the talus using a tendon retriever and secured with 5-0 chromic suture. The vascular anastomosis in all cases was achieved by using an operating microscope. End-to-end anastomosis between the anterior tibial artery and descending genicular artery was performed in an interrupted fashion using an 8-0 nylon suture. The tourniquet around the thigh was then released and blood flow to the flap confirmed using a doppler probe. The flap was then docked into the talus and secured using chromic suture. The wound was closed in layers in standard fashion, followed by placement of a sterile dressing and trilaminar splint.

Postoperatively, while patients were in the hospital, flaps were monitored using a handheld doppler. The pedicle of the flap was located subcutaneously and easily found using a doppler. Flaps were not actively monitored after patients left the hospital. Incidentally, on follow-up contrast-enhanced imaging, contrast could be seen in the flap, indicating blood flow to the flap was present. While the purpose of the imaging is to monitor the status of the AVN and not monitor the flap, imaging can be used to confirm the viability of the flap.

During the postoperative period, patients were kept nonweightbearing for 2 weeks. A wound check was performed at that time and the patient was subsequently transferred into a CAM walker boot. Patients remained nonweightbearing until 4 to 6 weeks postoperative but were allowed to perform active range of motion as tolerated. Weightbearing 3-view (anteroposterior, lateral, and mortise) x-rays of the ankle were performed. At this time, the patient was progressed to weightbearing as tolerated in the CAM boot and physical therapy was initiated. The patient was then weaned from the boot into a sneaker as tolerated over the subsequent 6 weeks.

Outcome Evaluation

At a minimum of 6 months postoperatively, patients were contacted via phone call to complete the Foot and Ankle Ability Measure–Activities of Daily Living (FAAM-ADL) subscale, ⁹ Short Form-12 (SF-12) mental composite scale (MCS) and physical composite scale (PCS), and visual analog scale (VAS) for pain (0-100). Patients were also asked about their satisfaction with the outcome of the surgery (very dissatisfied, dissatisfied, neither satisfied nor dissatisfied, satisfied, very satisfied) and likelihood to repeat the surgery (very low, low, moderate, high, very high) if given identical conditions using 5-point Likert scales. Patients were also asked about pain or other complications experienced at the harvest site following surgery.

Pre- and postoperative magnetic resonance imaging (MRI) without contrast was reviewed by a fellowship-trained musculoskeletal radiologist who was blinded to duration of when the MRI was obtained postoperatively and patient clinical outcomes. Images were read using the SECTRA (Linköping, Sweden) picture archiving and communication system (PACS).

Patient demographics and AVN characteristics were collected via electronic chart review. Descriptive statistics (means, standard deviations, ranges) were reported for patient demographics and outcome variables. Paired t tests were used to compare pre- and postoperative functional measures. Statistical significance was set at P < .05.

The etiologies of talus AVN included trauma (2/6, 33%), sepsis-related treatment (2/6, 33%), and idiopathic (2/6, 33%). All 6 lower extremities were in the precollapse stage of AVN. ¹ Preoperative MRI revealed varying degrees of AVN in the talus but was generally located throughout the entire talus (body, neck, and head) in each patient. A total of 6 MFC periosteal flaps were performed in 5 patients from 2016 to 2018 with an average age of 44.2 (range, 37-67) years, all of whom were females, with a mean body mass index (BMI) of 30.0 (range, 21.0-38.1) kg/m² (Table 1). Average duration of symptoms prior to surgery was 39.5 (range, 13-84) months.

OPEN IN VIEWER

Table 1.

Medial Femoral Condyle Periosteal Free Flap Patient Demographics.

Patient No.	Age, y	Sex	BMI, kg/m2	AVN Cause	Laterality	Smoking Status	DM?	Duration of Symptoms, mo
1	67	Female	30	Trauma	Right	Never	No	13
2a	38	Female	38	Sepsis	Right	Former	No	72
2b	39	Female	37	Sepsis	Left	Former	No	84
3	37	Female	21	Idiopathic	Right	Never	No	18
4	41	Female	26	Idiopathic	Left	Never	No	18
5	40	Female	28	Trauma	Left	Never	No	29

Abbreviations: AVN, avascular necrosis; BMI, body mass index; DM, diabetes mellitus.

Results

All 6 limbs achieved full weightbearing status and none of the patients required revision surgeries during the immediate follow-up period. At an average of 16.9 (range, 6-28) months postoperatively, patients reported a mean FAAM-ADL subscale score of 93.1 ± 8.3 (range, 77.5-98.8), which was a 63.2% improvement from the reported mean preoperative FAAM-ADL of 57.0 ± 14.8 (range, 44.1-83.3) (P = .004) (Table 2). As part of the FAAM survey, patient-reported single assessment numeric evaluation (SANE) of daily activity level of function also significantly improved by 88.1%, from a preoperative value of 44.2 ± 23.8 to 80.8 ± 11.1 postoperatively (P = .038). Also, as part of the FAAM questionnaire, 5 of 6 (83.3%) patients also reported their current function to be “normal” or “nearly normal” in the lower extremity that was operated on. SF-12 mental and physical health composite scales both showed no significant difference following surgery (Table 2). VAS for pain significantly improved by 77.6%, decreasing from a mean preoperative value of 8.17 to a postoperative value of 1.83 (P = .003).

OPEN IN VIEWER

Table 2.

Comparison of Mean ± SD Pre- and Postoperative Outcome Measures.

Outcome	Pre	Post	P Value
FAAM-ADL (/100)	57.0 ± 14.8	93.1 ± 8.3	.004 a
ADL-SANE (/100)	44.2 ± 23.8	80.8 ± 11.1	.038 a
SF-12 MCS	57.9 ± 6.6	52.3 ± 8.0	.241
SF-12 PCS	27.9 ± 6.3	37.8 ± 14.5	.150
VAS (/10)	8.17 ± 1.60	1.83 ± 2.23	.003 a

Abbreviations: ADL-SANE, activities of daily living–single assessment numeric evaluation; FAAM-ADL, Foot and Ankle Ability Measure–Activities of Daily Living; MCS, mental composite scale; PCS, physical composite scale; SF-12, Short Form-12; VAS, visual analog scale.

a

Indicates statistical significance (P ≤ .05).

In addition, 100% (5/5) of patients reported being “very satisfied” with the surgery and 5 of 5 (100%) reported either a “very high” or “high” likelihood they would repeat the surgery if given the same previous circumstances. Five of 6 limbs (83.3%) had no pain or discomfort at the medial femoral condyle harvest site, with 1 patient reporting minimal paresthesias. There were no major postoperative complications, such as wound dehiscence or thromboembolic events.

At a mean of 12.7 (range, 6-25) months postoperatively, ankle MRI without contrast was reviewed by a fellowship-trained musculoskeletal diagnostic radiologist. All 5 patients (100%) demonstrated resolution of marrow edema and soft tissue inflammation surrounding the areas of AVN when compared to preoperative imaging (Figure 5). In addition, AVN was not shown to progress in any patient.

OPEN IN VIEWER

Figure 5.

(A) Preoperative magnetic resonance imaging findings demonstrating avascular necrosis. (B) Six-month and (C) 12-month postoperative images demonstrating progressive improvement of marrow edema surrounding areas of avascular necrosis and decreased joint effusion.

Discussion

Treatment of avascular necrosis of the talus remains a challenging and controversial topic, with multiple operative and nonoperative options available. However, nonoperative measures, such as a trial of nonweightbearing, bracing, medical management, and bone stimulation, are often not effective, thereby necessitating a surgical procedure. Joint preservation techniques include core decompression, nonvascularized bone graft, allograft, or vascularized bone graft. Vascularized bone graft techniques offer the most promise but can be technically challenging with potentially significant donor site morbidity and relatively increased surgical time. This article presents a novel approach to the treatment of avascular necrosis of the talus. The periosteal free flap (pMFC) is a modification of the traditional medial femoral condyle free vascularized bone graft, which, based on these preliminary results, shows early success in treating AVN of the talus. The pMFC is a technically easier procedure offering decreased operative time, equivalent results, and less donor site morbidity, though no direct comparison of these factors was performed in the current investigation. The harvest involves only the periosteum and preserves the cortex. This enables the flap to be harvested without any risk of fracture to the femur. In addition, the periosteum is more pliable, which allows the flap to be placed in the talus without creating a large access site.

Many alternative surgical treatment options exist for this difficult problem. Percutaneous drilling for AVN in 1 study resulted in a 91% clinical success rate in 44 ankles from 31 patients. ¹⁰ Another technique using vascularized pedicle bone graft from the cuboid for treatment of talar AVN followed 13 patients for 2 to 12 years (mean, 6 years), comparing preoperative SF-12 and postoperative SF-36 scores, and showed significant improvement on the physical component summary by 23.3 ± 18.9 points and mental component summary improvement by 39.4 ± 10.1 points. ¹¹ In a retrospective review of MFC flaps used in the foot and ankle of patients who had failed nonvascularized bone graft procedures, at an average follow-up time of 20 months, all cases resulted in union. ⁷ Whereas Haddock et al ⁷ and Doi and Hattori ⁴ used a thick corticoperiosteal graft, our technique involves a less invasive approach with harvesting of only a thin layer of periosteum, with no known age limit of its use known to us. While many techniques describe the use of the ascending branch of the medial femoral condyle artery, we used the descending branch because it requires less dissection.^6,10

In this case series of 6 pMFC flaps in 5 patients, at an average postoperative follow-up of 16.9 (range, 6-28) months, no reoperations or major complications were reported, and all patients achieved full weightbearing status postoperatively. The mean FAAM-ADL subscale scores showed a 63.2% improvement, and SANE score increased by 88.1%. The operative extremity was graded as “normal” or “nearly normal” in 5 of 6 patients, and VAS scores decreased from a preoperative value of 8.17 to 1.83 postoperatively. Furthermore, the cohort’s mean SF-12 PCS score improved by nearly 10 points (37.8 vs 27.9), while postoperative mean MCS score was higher than the mean SF-12 MCS score in the United States for individuals aged 35 to 44 (52.3 vs 48.8) years. ¹⁴ The only noted complication was in 1 patient, who reported minimal paresthesias at the harvest site. In addition, postoperative MRI showed 100% resolution of marrow edema, without progression of AVN. The technique presented in this publication is based on the biologic potential of the periosteum alone, as several studies have provided insight into the role of periosteum in autograft growth, healing, and regeneration, likely secondary to the presence of multipotent mesenchymal stem cells. ¹⁵

One limitation of this study is the small sample size. As this is a relatively uncommon diagnosis, obtaining enough patients for a larger scale study can be difficult and is a shortcoming of the literature on AVN of the talus in general. In addition, the length of follow-up was relatively short, at a mean postoperative follow-up time period of 16.9 months. As this was a retrospective review, there is no direct comparison group to an alternate surgical technique. Finally, all patients were from out of state, and financial hardships prevented regular follow-up office visits with the operating surgeons. Physical exams were performed by their local orthopedic surgeon at scheduled follow-up visits, and therefore, we are lacking objective clinical measurements of postoperative function.

In conclusion, the pMFC flap is a viable alternative surgical option for treating patients with avascular necrosis of the talus when conservative treatments have not been successful. While the technique described does require the expertise of 2 surgical teams (orthopedic and plastic or microvascular surgery), the procedure is technically less challenging and requires less operative time than a traditional free vascularized bone graft, with the added benefit of decreasing donor site morbidity. In this small series of patients, the pMFC flap resulted in substantial pain relief and return to function. This study provides a valuable basis to initiate a prospective study with a larger sample size to improve the surgical treatment methods for patients with AVN of the talus.

Conclusion

The pMFC is a novel modification of a previously described vascularized bone graft. Modification to the pMFC appears to have similar results as compared to the traditional MFC flap and be safe and effective in the short term with excellent clinical and radiographic outcomes. Additional studies are necessary to evaluate this promising surgical technique in the management of this extremely challenging diagnosis.