PIPJ Arthrodesis with a Thermal-Activated Metallic Implant

Abstract

Hammertoe deformity is a common pathology seen by today’s foot and ankle surgeons. The deformity of the lesser digits is a frequently encountered pathology in foot and ankle clinics. Arthrodesis is accepted as the gold standard for hammertoe correction. Surgeons performed this by removing the adjacent articular surface of the proximal interphalangeal joint to allow osseous fusion. Soule first described the procedure in 1910, and Jones later modified it to utilize a more straightforward dorsal approach [6]. In 1940, Taylor added the use of a Kirschner wire fixation for stabilization [7]. This is a simple and effective method of arthrodesis which remains a steadfast most common technique and practice despite its anterior weaknesses [1–4]. With modern intramedullary fixation systems, the goal is to address or remove the concerns surgeons encounter with K-wire fixation. Specifically, in a study involving 149 toes and 99 patients, Richman and colleagues noted positive results for intramedullary devices over K-wires for hammertoe fixation [3]. In discussing proximal interphalangeal joint arthrodesis for hammertoe correction, Ellington indicated a preference for intramedullary devices in order to avoid complications with K-wires [5].

Hammertoe deformity is a common pathology seen by today’s foot and ankle surgeons. The deformity of the lesser digits is a frequently encountered pathology in foot and ankle clinics. Arthrodesis is accepted as the gold standard for hammertoe correction. Surgeons performed this by removing the adjacent articular surface of the proximal interphalangeal joint to allow osseous fusion. Soule first described the procedure in 1910, and Jones later modified it to utilize a more straightforward dorsal approach [1]. In 1940, Taylor added the use of a Kirschner wire fixation for stabilization [2]. This is a simple and effective method of arthrodesis which remains a steadfast most common technique and practice despite its anterior weaknesses [1,4,5,6]. With modern intramedullary fixation systems, the goal is to address or remove the concerns surgeons encounter with K-wire fixation. Specifically, in a study involving 149 toes and 99 patients, Richman and colleagues noted positive results for intramedullary devices over K-wires for hammertoe fixation [5]. In discussing proximal interphalangeal joint arthrodesis for hammertoe correction, Ellington indicated a preference for intramedullary devices in order to avoid complications with K-wires [7].

Innovative Hammertoe Memory Devices

Nickel titanium , also known as nitinol (part of shape-memory alloy), is a metal alloy of nickel and titanium. Nitinol alloys exhibit two closely related and unique properties: shape-memory effect (SME) and superelasticity (SE, also called pseudoelasticity (PE)). Shape memory is the ability of nitinol to undergo deformation at one temperature and then recover its original, undeformed shape upon heating above its “transformation temperature.” Superelasticity occurs at a narrow temperature range just above its transformation temperature; in this case, no heating is necessary to cause the undeformed shape to recover, and the material exhibits enormous elasticity, some 10–30 times that of ordinary metal. The word nitinol is derived from its composition and its place of discovery: Nickel-Titanium-Naval Ordnance Laboratory. William J. Buehler [8] along with Frederick Wang [9] discovered its properties during research at the Naval Ordnance Laboratory in 1959 [10, 11]. Buehler was attempting to make a better missile nose cone, which could resist fatigue, heat, and the force of impact.

Several varieties of a nitinol metal implants have been utilized as an intramedullary fixation device. These implants require specialized instrumentation for insertion, and certainly a learning curve exists before a surgeon is able to insert the device in a uniform and predictable manner and also in an efficient amount of time. These implants do allow the surgeon to select a device that will stabilize the arthrodesis perfectly straight or one that allows for a small degree of flexion, thus allowing for a more anatomic configuration to the toe.

Variations of the above discussed implants have also been presented. One- and two-piece designs are marketed that incorporate threaded or barbed ends. One design incorporates a threaded screw-like portion that is initially inserted within the proximal phalanx. Distally, this implant possesses a barbed spade-like configuration that pierces and maintains position within the middle phalanx. Certainly, the question comes up of whether compression is necessary for successful hammertoe repair.

Screws are devices which provide more predictable purchase and compression. However, the majority of single-piece constructs required an additional incision to the distal tuft for insertion. Two-piece screw-in devices avoided the additional incision but added to complexity, not to mention an increased fear of tissue handling such as small implant components. In addition, there may be possible disengagement of the two-piece implant with the complication and failure of union at the PIPJ.

Prior studies have identified nearly more than 70 different products available over the past few years with notable progressive improvement in material quality, design, and ease of use for the hammertoe implants [12]. Early generation products were simple, often composed of static metallic of or absorbable composites that acted as positional fixation. The disadvantages were the lack of pullout strength providing no compression and inability to conform to the naturally variable anatomy of the osseous proximal phalanxes prompting increased design improvement. Later, new generation devices incorporated threads or winged arms to provide additional stability and compression across the fusion site. The devices commonly required storage in a frozen state prior to direct application and implantation. This limitation posed inherent difficulties as it added an urgency for insertion once the device was exposed to room temperature and began early expansion, in addition to occasionally providing a freeze-dried state that required warmth activation (Fig. 9.1a,b).

Fig. 9.1

(a) The HammerLock II implant and 2 weeks with good alignment. (b) The Hammerlock implant with arthrodesis at 6 weeks with full return to shoes and activity

Amiot, R.A. and Nowak, J. in a retrospective review looked at the use of HammerLock internal implant for fixation of PIPJ arthrodesis in the correction of fixed hammertoe deformities of the lesser digits. The osseous union in 20 out of 21 toes was 95% with an overall ACFAS-modified forefoot score improvement. Successful arthrodesis was achieved with a stable, asymptomatic pseudoarthrosis in one patient overall and one toe. The conclusion was that overall, there was extremely high fusion rates and a statistically significant improvement in pain and function scores postoperatively with very low complication rates (Fig. 9.2a, b, c).

Fig. 9.2

(a) Preoperative anteroposterior projections of second, third and fourth hammertoe contractures. (b) Anteroposterior projection showing good alignment of memory nitinol intramedullary hammertoe correction involving toes 2, 3, 4. (c) Lateral projection demonstrating good sagittal plane alignment of memory nitinol intramedullary hammertoe correction involving toes 2, 3, 4

Sandu, J.S.; DeCarbo, T. W.; and Hofbauer, H.M., in a retrospective study showed overall 93.8% fixation rate was noted in 35 patients who underwent surgical correction with the use of 65 implants for correction of hammertoes. The arthrodesis of the PIPJ utilized a one-piece memory nitinol intramedullary fixation device in which 27 presented in the follow-up study. Complications were identified in four patients for a 6.1% finding, including one asymptomatic non-union (1.5%), two hardware failures (3%), and one implant displacement (1.5%). None of the patients required revisional surgery.

Colliard, Jean-Yves; Gianfranco, John Petri; Van Damme, Geert; Deprez, Patrick; and Laffenetre, Olivier in a prospective multicenter study used an angulated intramedullary fixation for PIPJ fusion with the use of the IPP-ON implant. This study by Integra included 117 patients where 156 had hammertoe digital corrections and concluded an 83.8% arthrodesis rate after 12 months. There were five post-op complications (3.2%), two cases were treated intraoperatively, and one case required revision.

What Advantages Do Intramedullary Hammertoe Devices Offer

Intramedullary devices have various benefits over the standard percutaneous Kirschner wire fixation [12].

Typically, the distal interphalangeal joint is not violated with the placement of most intramedullary systems in comparison with K-wires. Beyond sparing uninvolved joints, intramedullary devices provided multi-planar resistance to rotational forces with the majority of intramedullary implants providing a source of compression across the fusion site. This compression is lacking with the standard percutaneous Kirschner wire fixation.

A recent retrospect review of 136 s PIP joints in 114 patients found improved postoperative radiograph alignment with the use of intramedullary implants in comparison to traditional K-wire [12]. Addiotionally, the placement of internal fixation in the intramedullary canal neglects the risks of pin-tract infection associated with percutaneous exiting of the Kirschner wire fixation. Elimination of percutaneous fixation alleviates anxiety and complications associated with pin site care and the appearance of percutaneous pins including cosmesis. The placement of internal fixation also lessens the concern for traumatic disruption of exposed fixation. The use of modern memory nitinol implants often simplifies the postoperative course as opposed to Kirschner wire fixation. K-wire patients are likely to have pins extruding from their toes as opposed to recommended intramedullary fixation devices which pose less pain, increased osseous fixation, and improved clinical appearance. Ultimately, patient’s satisfaction is more important regardless of radiographic or clinical appearance, and high satisfaction rates have occurred with intramedullary fixation devices [4, 13].

In a retrospective study of 30 PIPJ fusions using nitinol in 10 patients with diabetic neuropathy, Roukis cited a 93% success fusion rate with the remaining patients having a stable non-union [4].

Important Considerations in Surgical Planning

Surgical planning for hammertoes is far more complex than what we have learned over the years, especially as residents. Once examination and testing are complete, one has to consider not only the hammertoe correction but the stability and alignment of the MPJ in all planes as well. It is also imperative to evaluate the plantar plate, especially if there is a positive drawer sign (positive Lachman’s maneuver). The clinical exam also calls for suspicion of interdigital neuromas between the second and third interspace. X-ray examination, as well as MRI or 3-D CTs, may be obtained to further assess the clinical planning and surgical correction of toes.

It is often necessary to explain the potential need for plantar plate repair, metatarsal osteotomy, hallux valgus correction, and different hammertoe correction options prior to surgery. If there are flexible hammertoes on adjacent digits, attention should be considered for surgical correction on these adjacent digits simultaneously, as arthrodesis for one digit may necessitate a substitution phenomenon of the other subsequent toes. It is also important to adjust the staging and categorizing of the flexibility of the toe. If one toe has more extensor substitution and flexor substitution exhibited, definitive arthrodesis should be performed. This allows for better positioning of the toes, less swelling, and the more long-term standardized outcome. Equally important is a thorough evaluation of the plantar plate. If the plantar plate rupture or injury does exist, repair it accordingly. If the plantar plate is not repaired, a resulting crooked toe or loss of purchase is noted often with a shift medially or laterally. Surgeons may perform the plantar plate repair unilaterally or completely. Finally one may shift the metatarsal slightly medial or lateral during the procedure to allow for some correction of the toe position with an osteotomy.

Fixation Method and Surgical Technique

We recommend performing associated procedures for any forefoot surgery prior to addressing the hammertoe deformity. This includes a bunionectomy, tailor’s bunionectomy, and possible neuroma excision. This also decrease the stress on the toe and possible MPJ correction during the other procedures. Furthermore, consider the MPJ positions and metatarsal parabola with regard to the position of corrected first and fifth metatarsals.

The memory implants usually come in a straight design as well as 10-degree angle. Both have been used with a good cosmetic and functional result. The angled design assumes a “flexed fusion” position which may be more cosmetically pleasing to some, rather than a toe which is too straight. The large and medium sizes are usually placed within the second digit and medium and small implants placed within the third and fourth toes. It is uncommon to fuse the fifth toe yet more commonly seen within the orthopedic community. This is often seen in neurologic conditions as well as revisions.

A standard dissection technique for the arthrodesis is identical to that of the arthroplasty with regard to joint exposure. Standard dissection is carried down to the extensor tendon. Next a transverse or Z-lengthening extensor tenotomy is performed, releasing the extensor wing and sling mechanism both medially and laterally and across the MPJ. The IPJ capsulotomy is performed slightly proximal to the dorsal apex of the joint. The soft tissues are freed from the head of the proximal phalanx and the base of the middle phalanx. The head of the proximal phalanx is then resected recommending distal to the surgical neck and proximal to the articular cartilage. The oscillating saw blade is then reversed, and resection of the base of the middle phalanx is performed. Care should be taken to remove any bony fragments or ledges in the depths of the proximal phalanx cortex or middle phalanx cortex with no prominence. The bone ends are then manually opposed to assess sagittal, frontal, and transverse plane correction with bony apposition. Any transverse or sagittal plane osseous deformities can be corrected with angled cuts using the oscillating saw and soft tissue balancing. Next assessment of soft tissue correction with Kelikian testing at the metatarsophalangeal joint should be performed to assess any contractures in a sagittal, transverse, or frontal plane position.

The metatarsal osteotomies should be performed in conjunction to the correction of the digit if there is abnormal metatarsal parabola with structural elongated second or third metatarsals (Fig. 9.3a-f). It is imperative to assess the integrity of the plantar plate clinically, prior to doing the surgical correction of the second digit. It may be helpful to utilize a MRI prior to addressing the correction of these toes, especially the second MPJ. Next the correct size implant is selected with the assistance of fluoroscopy, and the corresponding drill bit is used to create a pilot hole in the proximal and middle phalanges. The drill bit is advanced in a retrograde fashion from distal to proximal into the medullary canal of the proximal phalanx, allowing for 75% of the cutting flutes to pass into the canal. Secondly the same drill bit is then passed in an antegrade fashion from proximal to distal into the medullary canal of the middle phalanx, allowing for 25% of the cutting flutes to pass into the canal. The current HammerLock II kit also has automatic stops designed into the drill system, so proper depth is easily achieved for each side. While stabilizing each phalanx, the corresponding broach is then used to broach the proximal and middle phalanges. Care must be taken not to rotate the phalanx or misalign the broach during the process. Pay special attention to assessing the medial lateral translation of the middle phalanx on the proximal phalanx when broaching the middle phalanx. Ideally, the medial and lateral cortices of the adequate phalanges should align to avoid symptomatic bony prominences once the joint is fused with the internal implant. Be sure to leave the middle tab in place. The exposed proximal end of the implant is then inserted into the prepared medullary canal of the proximal phalanx and advanced proximally until the clamshell/tab assembly makes contact with the distal bony surface of the proximal phalanx. A mallet may be used to gently seat the implant fully into the proximal phalanx. The forceps were then removed, along with the clamshell, exposing the distal portion of the implant. Be sure to leave the middle plastic tab and placed during this portion of the procedure. The exposed distal portion of the implant was then inserted into the prepared medullary canal of the middle phalanx; advancing the implant onto contact is made with the plastic tab. The plastic tab was then removed, and the arthrodesis site is manually compressed axillary for approximately 1–2 min to allow body heat activation. Manual compression is removed, and the arthrodesis site was then assessed directly under fluoroscopy for both excellent bone anastomosis and implant position (Fig. 9.4a–f).

Fig. 9.3

(a) Presentation of classic hammer toes with structural elongated metatarsals 2,3 right foot. (b) Lateral projection of rigid contracture of toes. (c) Preoperative anteroposterior projection of hammertoes involving the second and third toes. (d) Preoperative lateral projection of second and third hammertoe contractures. (e) Postoperative X-rays at 4 weeks with good alignment HammerLock fixation and 2,3 met osteotomies. (f) Postoperative X-rays at 6 weeks with arthrodesis and with return to activities

Fig. 9.4

(a) Preparation of the proximal phalanx with dissection of the extensor tendon. (b) Resection of the proximal phalanx head with dissection of the extensor tendon. Intramedullary drill with accurate alignment plane. (c) Middle phalanx is drilled and prepped, intramedullary canal is centered in good alignment and planes. (d) Middle phalanx is broached in accurate plane. (e) Insertion of the proximal phalanx implant with remaining exposed distal portion of the implant. This is then introduced into the medullary canal of the middle phalanx, advancing it forward and then removal of the central tab and manually compressing for 1–2 minutes. (f) Toe with implant properly seated. Proximal interphalangeal joint arthrodesis using the HammerLock implant

In the earlier designs, the surgeon was required to hold the two phalanges together after implantation for 1–2 min to allow for the implant to warm up and then expansion of the device to occur before there was enough stability to release the toe. In patients with osteoporotic bone, this often took longer for stability to occur. With the newer generation of implants, there is a significant difference in perceived stability almost immediately upon phalangeal impaction and placement. It is a noticeable difference which seems to equate to better overall outcomes.

A C-arm (fluoroscopy) can be used to confirm that the component is properly engaged and the toe is well aligned. If there is slight gapping of the implant at the proximal interphalangeal joint, compression of the middle phalanx onto the proximal phalanx should be performed to improve placement position, making sure the arms are distracting along the internal cortex. Extensor tendon closure is now performed with simple interrupted 3–0 absorbable suture, and the skin is then closed in a simple interrupted-type fashion.

The nitinol implants provide proportional sizing for better fit, with straight and angled versions and multiple sizes. The second-generation Hammerlock II design provides for higher strength with more aggressive barbs providing solid fixation and flexible legs to minimize stress on cortical walls. Most implants provide fast and easy insertion systems in the self-contained sterilized packs.

How to Avoid Complications

As with all implants, there are apparent weaknesses and potential complications . The superficial postoperative infections are among the most common postoperative complication as Roukis determined in a study of intramedullary fixation with a one-piece nitinol implant [4]. Due to the intramedullary nature of these implants, removal of the implant is often a concern for surgeons. I have not observed this to be a practical concern clinically. When the bone is well healed and solid, there is never a need to remove the implant deep within the bone. The material is nickel/titanium and is extremely biocompatible. The reasons to remove the implant usually stem from painful post-op problems, such as infection or non-union. When this is the case, there is relative ease in separating the two phalanges from each other and removing the implant manually. In several cases where the arms have broken thru the phalangeal wall, these patients have been asymptomatic, and implant removal was not required. The PIP joint space still was able to heal. Another case was stabilized with a K-wire from distal and placed across the MTPJ to provide stability for 6 weeks.

Prior experience with some of the first-generation memory hammertoe implants shows a fairly narrow waist. This is the area between the proximal and distal arms and acts as a point of directed stress with range of motion and excessive movement. Often the waist is located right in the joint space and is vulnerable to weight-bearing stresses when bone healing is delayed. When the waist fatigues and breaks, there is resultant instability often leading to a fibrous malunion. At times, bone healing did occur yet often with a slight deviation. This is not always symptomatic. The advent of the second-generation implants has provided much improved stability which does not appear to fatigue and fail like the earlier designs. Just a minor increase in the diameter of the arms provides tremendous increase in overall stability as well as adding more metal arms. The waist in the newer designs has been shifted slightly proximal, so now the weakest portion of the implant is resting within the proximal phalanx and is not prone to early movement and metal failure.

When implant loosening occurs, one must really analyze the cause. Perhaps the implant was not the proper size or too small, allowing unnecessary motion due to lack of osseous purchase. Cortical disruption may be seen once surgical drilling, broaching, and impaction of the implant. These are also potential pitfalls, increasing the risk of failure. Often manual impaction with splinting in the office setting may suffice. However, if there is failure to form a solid union or satisfactory fibrous union in shoes, then a revision may be required. The benefits from using an expanded implant which extends into the middle or proximal phalanx within the canals account for better compression and minimal loss of bone. Although we do not often think of intramedullary devices as an ideal fixation in the case of a failed index K-wire, a study by Ellington noted that fixation revision with intramedullary devices may be the best [7].

At times, technical difficulty with intramedullary implants is the other common concern. Given the lack of osseous volume and surface of the proximal and middle phalanges, this may require more initial precision with drill holes and preparation of the canals, especially when there is osteopenia, or poor bone quality. Excessive osseous distraction with multiple guide pins and drills will result in a loss of implant purchase and the potential need to resort to traditional K-wire fixation placement. Sung and co-workers noted superior alignment postoperatively in patients with implants and attributed the alignment to the increased need for precision and accuracy when using implants over the traditional wire fixation [12] (Fig. 9.5a, b).

Fig. 9.5

(a) Extraosseous fixation of the proximal legs, due to inaccurate drilling and alignment with technique and an undersized implant. (b) Less than accurate alignment with the implant placed too proximal into the proximal phalanx and overdrilling

There are some limitations of the implants available. The short or stubby intermediate phalanx does not always allow for complete settling of the distal prongs of the implants to lie within the phalanx. This may cause DIPJ injury or pain with intra-articular placement. A narrow proximal phalanx is also a possible concern for placement and expansion. Soft bone may also be a concern as the expansive nature of the device upon placement and deployment may possibly break through the stiffness of the phalangeal wall.

Postoperative Care

Postoperatively, we have utilized memory nitinol intramedullary devices in over 500 procedures, and patients are weight-bearing as tolerated in a surgical shoe with a splint-type dressing along the surgical performed toes. We have found with increased stability, there is less pain using these implants which allows easier postoperative pain management recovery. The patient is evaluated at 1 week postoperatively for the first dressing change and again at 2–3 weeks to remove the sutures depending on their healing rate. The toes are then splinted with digital splinted devices or Coban taping techniques. These devices are usually worn for an additional 2–4 weeks. It is not uncommon to see arthrodesis at a faster rate with osseous fusion at 4–6 weeks which would allow a quicker recovery. If a metatarsal osteotomy or MPJ release is performed, patients are still in a surgical shoe for approximately 6 weeks followed by transition to an athletic shoe gear at 5–6 weeks postoperatively. Weight-bearing is performed as tolerated, and exercise is limited for approximately 6–8 weeks or when arthrodesis occurs clinically and radiographically (Fig. 9.6a-d).

Fig. 9.6

(a) Lateral view of the foot showing painful hammertoes of toes 2, 3, 4. (b) Dorsal view of the foot with painful hammertoes of toes 2, 3, 4. (c) Anteroposterior view showing postoperative correction of hammertoes 2, 3, 4. (d) Lateral view with adequate sagittal plane correction of hammertoes 2, 3, 4

Conclusion

Innovations in foot and ankle surgery are advancing quickly with intramedullary hammertoe implants. Although current research evidence supports effectiveness and superiority in healing, the implant costs are highly variable, and ultimate cost-benefit studies are needed.

There are low complication fusion rates in comparison to the traditional wire fixation. It is our opinion these fusion rates and effectiveness of the nitinol implants provide the static continuous compression to allow effective arthrodesis and alignment in all planes. The inherited benefits of these nitinol devices clearly demonstrate their efficacy in place of hammertoe corrective surgery. Larger prospective studies and randomized control trials may offer future concrete answers and accepted gold standards of these implants. Nitinol implant fixation will likely become the new gold standard in years to come in arthrodesis of digital surgery. Lastly, with cost being a large concern, implant cost is highly variable, and ultimate cost-benefit studies are needed.