Vascularized Bone Flaps for Carpal Reconstruction

Abstract

Within the field of hand and upper extremity surgery, reconstruction of the bony carpus remains a perplexing task and is a field undergoing rapid evolution. Among the eight bones of the carpus, the scaphoid and lunate are most frequently affected by traumatic and avascular processes which render their articular surfaces degenerated and painful. These conditions include scaphoid waist fracture, scaphoid proximal pole fracture, and Kienböck’s disease of the lunate. While traditional salvage operations with limited functional outcomes have historically been employed for management of these unsolved problems, advances in microsurgical understanding and capability are changing the treatment algorithm at our center. This paradigm shift centers in large part around the introduction of new techniques for vascularized bone and cartilage transfer for carpal reconstruction.

Vascularized Bone

Vascularized free tissue transfer is a fundamental technique in the armamentarium of a plastic and reconstructive surgeon. While the majority of free tissue transfers aim to bring muscle and or skin from areas of excess to areas of reconstructive need (i.e. in breast reconstruction), clinical situations occasionally arise wherein the transfer of vascularized bone and/or cartilage is required. The most common indication for vascularized bone transfer is mandibular reconstruction after traumatic injury or tumor extirpation. The fibula free flap has long been used for this purpose and others, and this is a reliable tool for reconstruction of large bones. ¹

Though it is well understood that many challenges limiting the success of carpal reconstruction are related to insufficient bony vascularity, traditional bone flaps liked the fibula are far too large and potentially morbid to be useful for carpal reconstruction. Pedicled bone flaps based upon small vessels supplying segments of the distal radius have been applied to the task of carpal reconstruction with some success.² However, these are limited with respect to reach, size, structural integrity, reliability of blood supply. Furthermore, these local bone flaps cannot provide vascularized articular cartilage which is a key component that will be discussed in detail below.

Recently, a novel class of small vascularized bone free flaps has been introduced. This dyad of bone and osteochondral flaps are based upon the descending geniculate vessels which supply the medial femoral condyle and medial femoral trochlea. Utilization of this versatile donor site for vascularized bone free flaps has resulted in expansion of our ability to bring vascularized bone and cartilage to the carpus for the purpose of articular and extra-articular bony reconstruction. This advanced microsurgical approach to reconstruction of the carpus has revolutionized the treatment of challenging scaphoid fractures and Kienböck’s disease. Is the objective of this article to outline these recent advances and current treatment options.

Scaphoid Waist Non-Union

The scaphoid is the most commonly fractured bone in the carpus and the waist is the most common anatomic location for fracture to occur. In many cases, non-operative management in a cast or prompt open reduction and internal fixation leads to fracture union and favorable patient outcomes. However not infrequently, scaphoid waist fractures fail to heal after operative or non-operative management (Figure 1). Left untreated, scaphoid waist non-union may lead to pain and rapidly progressive arthritis and loss of wrist function. Risk factors for this unfavorable outcome include fracture displacement, delayed treatment, and smoking.³

Figure 1.

Scaphoid waist non-union visualized on a CT scan.

The traditional first line of treatment for scaphoid waist non-union is open reduction and infernal fixation of the fracture (which represent a second surgery for patients who were initially treated operatively) with non-vascularized cancellous or corticocancellous bone grafting.⁴ In many cases, this is successful, but still there are some patients who again go on to non-union. Recalcitrant scaphoid waist non-union despite sound surgical treatment may be multifactorial, but current thinking is that reduced vascularity of the proximal pole leads to difficulty with incorporation of non-vascularized bone graft. For this reason, vascularized bone free flaps, which bring new and exuberant blood supply to the fracture site, have gained in popularity.

While different sources of vascularized bone have been described for the treatment of scaphoid waist non-unions, we favor the medial femoral condyle donor site. Through a modest incision on the medial side of the knee, we are able to access the medial femoral condyle without disrupting the joint or nearby musculotendinous units. The dominant periosteal vascular supply to the medial femoral condyle is through the descending geniculate artery, which is dissected back to its origin from the superficial femoral artery (Figure 2). After isolating this vascular pedicle, its branches are followed onto the medial femoral condyle, where a small portion of bone supplied by these vessels is harvested using an osteotome.⁵ While large amounts of bone are available for harvest,⁶ a small 1cm × 1cm × 1 cm cube of vascularized bone is sufficient for treatment of the scaphoid waist (Figure 3). This bone and its vascular pedicle are brought from the knee to the wrist where the bone is used to fill the non-articular defect in the scaphoid waist, and the artery and vein are microsurgically anastamosed to the radial vessels to provide bony perfusion.

Figure 2.

The descending geniculate artery can be seen running parallel to the femur, and then climbing up onto the medial femoral condyle where it provides blood flow to local bone and cartilage.

Figure 3.

A medial femoral condyle bone flap has been harvested based upon its descending geniculate pedicle, and is ready for inset into the scaphoid and vascular anastomosis.

A growing body of literature suggests that MFC bone flap reconstruction of recalcitrant scaphoid non-unions has a high rate of success⁷ even in cases which have been refractory to other treatment. Clinical union is expected approximately 8–12 weeks after surgery, and after this point patients undergo a course of therapy focused upon regaining range of motion. Perhaps even more importantly, clinical studies have found that morbidity related to the knee donor site is exceptionally uncommon. Most patients return to unrestricted ambulation on the first post-operative day.⁸

Proximal Pole Scaphoid Non-Union

When scaphoid fractures occur within the proximal pole which abuts the radius, the reconstructive challenge is even greater. Because of the scaphoid’s retrograde blood supply, fractures in this location are at an elevated risk for non-union compared to waist fractures. In cases of extreme hypovascularity, the ischemic proximal pole not only fails to heal, but may undergo avascular necrosis and fragmentation (Figure 4). While this avascular segment indeed represents only a small portion of the scaphoid, it includes the critical articular surface which is transfers loads across the wrist and onto the radial platform.

Figure 4.

Proximal pole scaphoid non-union after an unsuccessful attempt at primary fixation with a headless compression screw. The fragment is very small and is not suitable for salvage. Instead, proximal pole reconstruction with a medial femoral trochlea free flap is indicated.

Small, fragmented, and/or necrotic proximal pole fractures are particularly problematic because in many cases the proximal fragment is unsalvageable. Because the MFC bone flap does not carry the cartilage necessary for proximal pole reconstruction, it is not a suitable solution for this fracture pattern. Fortunately, in addition to supplying the medial femoral condyle, the descending geniculate vessels also send a transverse branch to the medial femoral trochlea (MFT). The trochlea is the superomedial extent of the patellofemoral articular surface, and it has been shown to have minimal involvement in weight bearing. This articular segment can be harvested as a small osteochondral free flap which is very similar in shape to the native scaphoid proximal pole (Figure 5). Rather than bridging two fractured segments of the scaphoid at the waist (like the MFC), this MFT free flap is used to completely replace the unsalvageable proximal pole with healthy vascularized bone and cartilage (Figure 6).⁹

Figure 5.

The medial femoral trochlea osteochondral free flap can be used for reconstruction of the proximal scaphoid or lunate.

Figure 6.

The proximal scaphoid has been replaced by a medial femoral trochlea osteochondral free flap which is fixated to the distal fragment using a headless compression screw.

Results of MFT free flap reconstruction of the scaphoid proximal pole demonstrate reliable radiographic union and excellent objective and subjective patient reported upper and lower extremity function.¹⁰ Using this reconstructive technique, we are now able to reconstruct the scaphoid and preserve normal carpal anatomy in even the most challenging of fracture patterns.

Kienböck’s Disease

Kienböck’s disease is an idiopathic avascular necrosis of the proximal portion of the lunate which results in avascular necrosis and ultimately articular collapse (Figure 7). While we don’t know exactly why Kienbocks’ disease occurs, it has been observed that this condition is more common in patients with ulnar negative variance. This is likely due to a concentration of the carpal biomechanical load upon the radiocarpal joint, with little contribution of the ulnocarpal articulation.

Figure 7.

Coronal plate MRI of this patient with advances Kienbock’s Disease demonstrates an avascular lunate with collapse.

A variety of carpus-preserving techniques have been described for the treatment of early Kienböck’s disease prior to collapse of the lunate’s proximal articular surface. However, once the proximal portion of the lunate has collapsed, wrist salvage with proximal row carpectomy or limited intercarpal fusion has previously been indicated. While these operations perform reasonably well in older and lower demand patients, they may fail to offer the function and durability desired in a younger and/or higher demand patient.^11–14

As an alternative to wrist salvage in these high demand patients with proximal lunate articular collapse, the MFT flap (originally described for scaphoid reconstruction) has been adapted for proximal lunate reconstruction.¹⁵ Rather than excise the damaged lunate, its distal articular surface abutting the capitate is preserved and the proximal segment is replaced with vascularized bone and cartilage which closely mimics the shape of the native lunate (Figure 8).

Figure 8.

CT Scan of the wrist demonstrating a reconstructed lunate bone after medial femoral trochlea flap transfer.

Like the scenarios described above, survival and union of this transferred osteochondral segment is routine. This reliability is attributable to the microvascular anastomosis, which provides blood flow to the transferred tissue. While this procedure is still relatively new, medium term objective and subjective patient reported outcomes have been encouraging,¹⁶ and donor site morbidity has been found to be low.¹⁷

Sharpening the Cutting Edge: An Unconventional Collaboration to Advance Carpal Articular Reconstruction

Given the relatively novelty of vascularized osteochondral free flap reconstruction of the carpus, articular reconstruction of the proximal scaphoid and lunate has become the focus of a successful interdisciplinary research effort at our institution. The central question being addressed in this endeavor is exactly how well the MFT osteochondral free flap mimics the native scaphoid and lunate articular surfaces it is meant to reconstruct.

Previous surgical studies of this issue have used rudimentary 2-dimensional methods to describe intersurface correspondence.^18,19 Because the carpal and femoral articular surfaces are complex 3-dimensional surfaces, these methods are insufficient and a new approach is warranted. To this end, our group (led by myself and Dr. Amelia Van Handel) has leveraged the incredible range of expertise residing on the Washington University campus and formed a collaboration with Drs. Kari Allen and Leigha Lynch from the department of anatomy and neuroscience. While these researchers usually focus on the evolution of primate brain size and skull shape using medical imaging and 3-dimensional morphometric methods, we found that their expertise in quantities 3D modeling of complex surfaces could be applied to carpal articular reconstruction. Over the course of two years, this team of unlikely collaborators has developed a protocol for quantitative assessment of different surgical options for carpal articular reconstruction (Figure 9). At this point we are able to use MRI data to quantitatively assess carpal and femoral articular morphology to assess and compare the suitability of an individual’s anatomy for MFT or other reconstructive techniques. In the future, we hope to validate and apply this technique to provide precision and individualized patient counseling and surgical planning.

Figure 9.

The proximal capitate (A) and the MFT flap (B) have been digitally overlaid upon the native lunate to generate a heat map of local intersurface distances. Given the more favorable correspondence seen between the MFT and lunate, MFT reconstruction may offer a morphologic advantage over proximal row carpectomy in this patient.

Conclusion

Recalcitrant scaphoid waist fractures, proximal pole scaphoid fractures with an unsalvageable articular fragment, and advanced Kienböck’s disease in the young patient are among the most difficult problems facing hand surgeons today. Recent advances in the field of vascularized bone transfer have expanded our treatment options such the carpal preservation is now available to most patients. Specifically, the MFC and MFT free flaps based upon the descending geniculate vessels provide small, flexible, and expandable sources of vascularized bone and cartilage which are suitable for use in the bony carpus. While these procedures are relatively new, mid-term outcomes and our own institutional experience have been encouraging.