The decision to replace or fuse the first MTPJ is an age-old argument among foot and ankle surgeons. Proponents of arthrodesis maintain that fusion provides excellent pain relief and a stable first ray that is functional in nearly all aspects of average daily ambulatory function and that the results are long lasting with low revision rates. Proponents of joint replacement cite preservation of ROM as the primary and necessary goal when treating first MTPJ pain and arthrosis. However, the long-term durability and complication rates with implants have been discussed by many researchers [9, 10, 34]. Both procedures are joint destructive, so an appropriate differentiating description of implant arthroplasty versus arthrodesis would be motion preserving versus motion sacrificing. Inherent in this discussion of motion-preserving and motion-sacrificing procedures is the benefit that each class of procedure provides for the patient with painful first MTPJ deformity and arthrosis. Both procedures provide a decreased level of pain compared to preoperative values in short- and midterm patient series. The question at the forefront of the argument is whether it is restoration of motion that produces positive outcomes or whether it is simply pain relief resulting from resection of the diseased joint during both arthrodesis and implant arthroplasty.

If restoration of range of motion (ROM) is the mechanism by which implant arthroplasty restores function and results in pain relief and satisfaction, one would expect outcome studies to highlight robust postoperative ROM measurements as a long-term benefit of the procedure as is noted in hip and knee replacements. With respect to pain relief and patient function, review of the favorable outcomes following first MTPJ arthrodesis seems to contradict the premise that it is improved ROM producing the positive outcomes in implant arthroplasty and support the idea that these results are achieved simply by removing the arthritic joint, thereby allowing the patient to ambulate and function better with less pain. With this dichotomy in mind, it is helpful to understand the real differences and similarities between replacement and arthrodesis. We undertook a detailed review of the published literature for first MTPJ implant arthroplasty to specifically quantify expected first MTPJ ROM values following replacement arthroplasty and therefore better understand the variables that drive positive patient outcomes.

We performed a systematic review of first MTPJ ROM after implant arthroplasty (Dayton et al. unpublished work). A total of 90 studies were identified in a systematic search of the databases. After the abstract review and application of the inclusion criteria, we identified 35 studies, 22 prospective studies that reported total preoperative and postoperative ROM without specifying dorsiflexion and plantarflexion ROM, and 13 studies that reported both dorsiflexion and plantarflexion ROM individually. The demographic information was not consistently reported among the included studies and therefore could not be analyzed. From the studies included, the mean follow-up duration was 44.7 months. For analysis, we divided the studies into two sets, one that reported only total preoperative and postoperative ROM without specifying dorsiflexion and plantarflexion ROM and a second set of the studies that specified the components of ROM. The results are listed in Tables 14.3 and 14.4.

Our review for MTPJ implant arthroplasty found that little consistency exists between studies. Reports of subjective satisfaction with decreased pain after joint replacement are undeniable and comparable to arthrodesis . However, one must ask if this finding is the product of the preservation of joint motion or if this finding is merely a result of the removal of a painful pathologic joint. The data of this review would point to the latter as there appears to be a relatively low net gain in first MTPJ ROM after implant arthroplasty, consistent with what Gibson and Thomson [34] noted in their comparative series of arthrodesis versus implant arthroplasty. These findings contradict the premise that there is robust motion after replacement; in fact, these average postoperative values fall into stages of hallux limitus/rigidus based on some classifications [17]. The end ROM measurements and hallux positions are also arguably not significantly different from the standard position of the hallux after a first MTPJ arthrodesis. Typically, the hallux is positioned in 10–15° of dorsiflexion relative to the first metatarsal axis. As Fig. 14.2 depicts, when this position is combined with the average first metatarsal declination angle of 20–25°, this provides a functional platform of approximately 30–40° of first MTPJ dorsiflexion relative to the first metatarsal declination . This is functionally the same position the hallux would be in during push-off following an implant arthroplasty that produces 35–45° of dorsiflexion. The fused hallux is in a functional position in the sagittal plane and maintains weight-bearing function despite the loss of range of motion. This functional position of the hallux is similar to the average dorsiflexion of 42° achieved in normal gait as identified by Nawoczenski et al. [65].

The approach to a first MTPJ arthrodesis begins with a skin incision located dorsally, just medial to the extensor hallucis longus (EHL) tendon and central to the long axis of the first metatarsal and proximal phalanx as shown in Fig. 14.3. The length of the incision will vary based on the hardware utilized for fixation but generally starts just proximal to the neck of the first metatarsal and extends to a point just proximal to the head of the proximal phalanx. The incision is deepened through the subcutaneous layers directly without wide subcutaneous undermining, taking care to avoid damaging neurovascular structures and maintaining full-thickness soft tissue flaps medial and lateral. Once the dissection is completed through the subcutaneous layers, an incision through the periosteum and dorsal joint capsule begins the development of the surgical pocket. The periosteum and joint capsule are carefully elevated from the distal metatarsal shaft, the head of the first metatarsal, and the base and proximal aspect of the proximal phalanx along the full extent of the incision. The osseous resection is carried out within this intracapsular and subperiosteal pocket with the neurovascular structures protected and maintained within the medial and lateral full-thickness flaps .

Joint resection should be aggressive and remove all osteophytes, cartilage, and subchondral bone from the entire head of the first metatarsal, including the plantar aspect, and all cartilage and subchondral bone from the entire base of the proximal phalanx. There are often degenerative changes and fibrosis present within the sesamoid complex, making it imperative to complete a plantar sesamoid release relieving any lateral contracture or ankylosis to allow the metatarsal to move laterally and the IMA to reduce. Complete release of the soft tissue attachments to the metatarsal head will also allow appropriate positioning of the proximal phalanx in all three planes. The plantar sesamoid release can be completed bluntly with a large key or similar elevator between the plantar metatarsal head and the superior surface of the sesamoids until the hallux is able to move freely and there is no longer any ankylosis remaining. Removal of the plantar contours of the head, including the crista, also acts to decompress the joint plantarly and reduces the chances of plantar pain postoperatively. We do not specifically prepare the sesamoids for fusion to the metatarsal; rather this is to decrease the cubic content of prominent bone plantarly.

Preparation of the bone surfaces can be completed by several methods, including rongeur and burr, specialized cup and cone reamers, or planar cuts. Planar cuts tend to be utilized for revision surgery rather than for a primary fusion procedure and do not provide as much bone-to-bone surface area for healing or as much fine-tuning ability, making satisfactory positioning more difficult to achieve. The cup and cone preparation allows robust maneuverability and adjustability for appropriate positioning in all planes. The authors’ preparation of choice is manual resection of the cartilage and all subchondral bone on the metatarsal head with a rongeur and preparation of the phalanx base with a round or oval high-speed burr as noted in Fig. 14.4. This allows the surgeon complete control of the bone shape and allows efficient and adequate removal of all subchondral and peripheral bone. Issues we have seen with power reamers include failure to resect the subchondral plate completely and fracture of the bone segments, especially in osteoporotic bone and when degenerative cysts are present .

To promote robust healing , we strongly recommend removal of all cartilage and the entire subchondral plate from the first metatarsal head and phalanx, down to bleeding trabecular bone. As the head is being prepared, the shape and size will be predicated on the size of the proximal phalanx base. The peripheral cortex of the base is left intact for stability, and the central portion of the base is deepened to sit on the head of the first metatarsal in a cup and cone (ball-and-socket) fashion. Care should be taken to keep the walls of the proximal phalanx cup steep and to keep the depth of the site equal throughout, avoiding any areas of varying depth which could limit good bone-to-bone apposition and cause instability .

Once the opposing surfaces are adequately prepared and a stable fit between them is present, the area should be irrigated and care taken to ensure there is no soft tissue within the arthrodesis site. Temporary fixation of the arthrodesis site is then performed. The authors’ temporary fixation is performed with a 0.062 k-wire in a manner similar to a hammertoe with the pin started in the central base of the proximal phalanx and driven distally out the tip of the hallux. The hallux is then positioned onto the first metatarsal head with appropriate hallux position in all three planes, and the pin is driven proximally into the first metatarsal head and shaft, as shown in Fig. 14.5a. With this technique, if repositioning is necessary, the pin can be pulled back out of the metatarsal and then advanced again after repositioning without the need to fully remove the temporary pin.

Positioning is evaluated clinically and radiographically with the temporary fixation in place. Key components of the position are the transverse , sagittal, and frontal plane position of the hallux. The hallux nail should be facing straight up to ensure the appropriate frontal plane position. Residual valgus rotation can cause irritation at the medial condyle of the proximal phalanx. The transverse plane position should be approximately 10–15° abducted relative to the first metatarsal shaft and sit in a well-aligned position relative to the second digit. There are several nuances that should be considered with respect to the transverse plane positioning of the hallux. If there is a high hallux interphalangeal angle (HIA) , the proximal phalanx may need to be positioned with less abduction than is normal to ensure that the distal tuft of the hallux doesn’t irritate the second digit.

The sagittal plane position is evaluated with a flat surface placed against the entire plantar surface of the foot and loading of the forefoot with the rearfoot in neutral position. The authors have found the best result with the proximal phalanx head positioned just barely off the surface of the plate with the rearfoot in neutral and the forefoot loaded. This is verified by a scalpel handle or the end of a forceps being able to just slide in under the distal aspect of the proximal phalanx, as shown in Fig. 14.5b. When evaluating position, one must identify if the hallux interphalangeal joint (IPJ) is pinned in a dorsiflexed or plantarflexed position and account for that.

The positioning in all three planes and the apposition at the arthrodesis site is first confirmed clinically, modified until appropriate, and then confirmed with intraoperative c-arm. Figure 14.6 shows a hallux that is positioned too high, with the head of the proximal phalanx well off of the weight-bearing surface. In this case, the pin wound be pulled out of the first metatarsal and the hallux repositioned in a lower position until it was just barely off of the surface as seen in Fig. 14.5. Once the positioning is confirmed, the site can be fixated. The authors’ fixation of choice is biplanar locking plates; however there are many different fixation options that have been described in the literature. Many of these fixation options and the subsequent healing rates associated with each are presented later in this chapter. Biplanar plating , shown in Fig. 14.7, allows the surgeon to easily manipulate the plates to fit the patient and the arthrodesis site, rather than having to fit the site to the plate. Biplanar plating has been shown to be stronger than certain anatomic plates [22] and is a more biologic form of fixation, providing a stable but not rigid fixation construct and allowing for micromotion to promote healing [67]. Biplanar fixation allows the surgeon to control all planes of motion and creates a very stable construct to aid in healing. Once the site is fixated in an acceptable position, the incision is closed in a layered fashion, and a dry sterile bandage is applied along with a light compression wrap.

Successful clinical outcomes of a first MTPJ arthrodesis are not only related to a successful union but also the union position, which affects overall foot function. If poorly positioned, additional pathologies and symptoms may result, potentially requiring additional revision surgery [35]. In a systematic review of union rate, Roukis [73] found a malunion rate of 6.1% (with 87.1% dorsal malunion and the remainder valgus rotation) and hardware removal rate of 8.5%. The author concluded that while the union rate was good, the malunion and hardware removal rates are inappropriately high and need further investigation for improvement. These results highlight the importance of appropriate fusion position and hardware selection and placement.

There are variations of recommended position in the literature; however the consensus appears to be approximately 10–15° abducted and between 10° and 25° dorsiflexed relative to the weight-bearing surface [14, 35, 46]. If the proximal phalanx dorsiflexion angle is too high, there is a resultant flexion contracture of the hallux IPJ and inability of the hallux to purchase the ground as noted in Fig. 14.8. The authors have found that excessive hallux dorsiflexion with the arthrodesis can result in severe pain to the plantar first metatarsal head and sesamoid complex. This results from increased weight-bearing pressure to the plantar first metatarsal head from the non-purchasing hallux and can lead to subsequent callus formation. Additionally, too much dorsiflexion is cosmetically unappealing, can also lead to hallux IPJ arthrosis form the contracture, and may result in pain to the distal hallux and shoe gear irritation from the dorsal hallux rubbing on the shoe. Similar to other authors [79], we have found the best position to be that noted previously in the description of the procedure, with the hallux positioned just barely off of the weight-bearing surface.

Alentorn-Geli et al. [2] have also reported that the dorsiflexion angle of the hallux post-arthrodesis plays a significant role in the pressure sub-first metatarsal head and hallux during gait. They reported that patients with a clinical dorsiflexion angle >15° and radiographic dorsiflexion angle >30° had more pressure under the first metatarsal head and those with a clinical dorsiflexion angle <15° patients had more pressure under the hallux. They recommended that in order to prevent increased plantar pressures under first metatarsal head, the arthrodesis dorsiflexion angle should be below 15° clinically and 30° on radiographs. This position allows the hallux IPJ room to plantarflex just slightly to contact the ground, without producing a flexion contracture. By using the scalpel or forceps handle technique previously described, the surgeon is also able to prevent the hallux from being positioned too low, which can result in dorsal jamming of the hallux IPJ with subsequent arthrosis. Additionally, if the hallux is positioned too low, there is a potential for lateral weight transfer during forefoot push-off, resulting in metatarsalgia.

The selected fixation construct can also play a role in malposition of the hallux. DeOrio [27] highlighted potential reasons for excessive dorsiflexion, including the use of pre-bent plates, and due to the anatomic configuration of the proximal phalanx. The author noted that a plate with no bend is actually able to provide approximately 15° of hallux dorsiflexion at the first MTPJ due to the conical configuration of the proximal phalanx. Since the plantar and dorsal surfaces of the proximal phalanx are not parallel but rather form an angle of approximately 30° to each other, a flat plate positioned over the joint after removal of the dorsal prominences and joint surfaces is able to form a functional dorsiflexion position. Therefore, as the bend of the plate increases, the dorsiflexion angle of the hallux will increase beyond 15°, potentially resulting in too high of a dorsiflexion angle. We have found this phenomenon to be common with pre-contoured locking plates and recommend removing the dorsiflexory bend from the plate prior to application to prevent hallux malposition. It is necessary with many of the available plates to bend the plate past 180° to fit over the convex contour of the dorsal first MTPJ and prevent excessive dorsiflexion position of the hallux. Similarly, Marsland et al. [60] found that the dorsiflexion angle of the first metatarsal and hallux measured by the dorsal cortices was significantly smaller than the intramedullary angle, with the dorsal cortices angle being 10.8° smaller. They concluded that since a pre-contoured dorsal plate for first MTPJ fusion uses dorsal cortices for the dorsiflexion angle, it may lead to excessive clinical dorsiflexion. Lewis et al. [51] evaluated the effect of plate position and type on the dorsiflexion angle in cadaveric specimens. They found that the more proximal the plate was placed, regardless of type, the greater the resultant dorsiflexion angle. This resultant change in angle was greatest for the pre-contoured plates, which highlights the importance of proper hardware selection and positioning for obtaining the ideal fusion position. Using the biplane construct described above, the plates do not participate in positioning of the site but are simply contoured to the fusion site after satisfactory position is achieved, thus avoiding the pitfalls of many “anatomic” plates.

The authors’ postoperative protocol consists of immediate protected weight bearing in either a surgical shoe or removable cast boot. The patient is usually encouraged to use an assistive device for the first 2–4 days postoperatively for balance and pain control. As tolerated, the patient may discontinue the use of the assistive device and bear full weight to the heel in the shoe or boot. The dressing is left in place until the first postoperative visit, typically 3–4 days after surgery. At that time, the dressing is removed, and the patient is instructed that they may get the surgical site wet in a shower [21, 32, 33]. A light dressing may be employed if needed, with compression from either an ace wrap or a compression stocking. The patient is encouraged to perform ankle joint range of motion without resistance daily. Radiographs are typically taken at 3 and 6 weeks postoperatively, and if clinical and radiographic signs of healing are present at 6 weeks, the patient is transitioned into a stable athletic shoe with flat-footed walking for 1–2 weeks, followed by gradual return to heel-to-toe walking at 6–8 weeks postoperatively. The average patient is released to all activity between 10 and 12 weeks .