Abstract
Errors in management of flatfoot fall into two broad categories: errors of decision making and technical errors related to surgery. Complications resulting from such errors can often be effectively managed with appropriate revision surgery. Other complications may occur as part of the natural history of the disease process, including arthritis and deformity. This chapter considers some of the more commonly encountered complications in this clinical setting, with a range of examples using the staging system for acquired flatfoot in adults.
Key Words
Revision, flatfoot, posterior tibial tendon, nonunion, triple arthrodesis, valgus
Errors in management of flatfoot fall into two broad categories: errors of decision making and technical errors related to surgery. Complications resulting from such errors can often be effectively managed with appropriate revision surgery. Other complications may occur as part of the natural history of the disease process, including arthritis and deformity. This chapter considers some of the more commonly encountered complications in this clinical setting, with a range of examples using the staging system for acquired flatfoot in adults, as described in the previous chapter. In the early 1980s it was still believed that one could repair the ruptured posterior tibial tendon (PTT), whether by end-to-end suture or with a Z-type shortening; however, it was not recognized that the rupture of the PTT represents only one aspect of numerous deformities that occur with the adult-acquired flatfoot deformity, and of course these attempted repairs were never going to work. Perhaps if this type of approach were used today by obtaining complete muscle and skeletal balance of the foot, it may be a treatment to consider.
In the 1980s, quite apart from repair of the PTT, many surgeons performed a flexor digitorum longus (FDL) transfer or a tenodesis of the FDL to the ruptured PTT. This does not make anatomic sense, because the ruptured PTT has an excursion of 12 mm, and a normal FDL tendon with an excursion of almost 2 cm, and it becomes very difficult to balance the force of these tendons. Why would one want to use a normal FDL and a tenodesis of tendon to a ruptured PTT? This was, however, the standard treatment used in the early 1980s. The use of the FDL continued as a stand-alone procedure continued in the mid-1980s, and authors believed that the use of the FDL alone would decrease pain of the ruptured tendon. One has to consider the cause of pain in the PTT rupture. Why is there pain, why does it occur, and is this always in the location of the tendon tear? Regardless of what in fact are the metabolic factors that cause these pain “generators,” they certainly disappeared in many patients with excision of the torn PTT and substitution with the FDL. This was the continued impetus to the treatment with an isolated FDL transfer. However, this too does not make any sense. So for almost a decade from the late 1970s through the mid-1980s the use of a transfer of the FDL into the navicular with excision of the ruptured PTT was the most common procedure performed in the absence of any additional procedures, such as osteotomies or limited arthrodesis. We recognize today that this would never work, and yet the concept of FDL transfer continued through today in one form or another. There have been numerous studies indicating that the results of this treatment—that is, FDL transfer subsequently combined with a calcaneus osteotomy—was satisfactory, but we now recognize that a transfer of the FDL is not the ideal procedure for correction of the muscle imbalance created by the deformity. The FDL is far weaker than the PTT, and regardless of the additional procedures performed to improve the structure of the foot, the muscle imbalance remains. Therefore we have to do something else to increase inversion power or at least weaken the eversion power. The peroneus brevis acts as a deforming force in the flatfoot deformity, and the unopposed pull of the peroneus brevis causes eventual elongation of the medial supporting structures, eventually leading to an abduction deformity. In the absence of the PTT, the functioning peroneus brevis therefore will always contribute to the worsening of the flatfoot deformity.
The isolated transfer of the FDL ignored the forces on the hindfoot as a result of the rupture of the PTT, and in the mid- 1980s, Myerson introduced the concept of adding a calcaneus osteotomy to the FDL transfer for management of the flexible flatfoot. Although Koutsougiannis in 1971 had recommended a calcaneus osteotomy for all flatfeet regardless of the underlying pathology or cause, this was not included in the PTT reconstruction until introduced by Myerson as a routine addition to the FDL transfer.
Although this approach was an improvement in the management of the flatfoot, it was still quite inadequate because it failed to recognize the many variations of the type of flatfoot deformities. We published our long-term results of this procedure in 2004 with reasonable outcomes, but began to note that in this group of patients for whom an FDL transfer and a calcaneus osteotomy was performed, many of them began to fail with recurrence of the flatfoot. This was largely the result of persistence of the muscle imbalance and the presence of deformity of the medial column, which had not been routinely corrected in this group of patients. This brings us to the present day where one now has to recognize the variety of deformities that exist in the adult flexible flatfoot deformity, in particular the various joints on the medial side of the foot that can sag or develop arthritis. The same principle applies to the various types of abduction of the forefoot or midfoot where the apex could be either at the talonavicular (TN), naviculocuneiform (NC), or tarsometatarsal (TMT) joint. Because of the plethora of these surgical alternatives, choosing a procedure can still be confusing. Decision making does, of course, depend on the severity of the deformity, the appearance of the foot, and the flexibility of the hindfoot and forefoot.
Perhaps the most important aspect of decision making is the presence of flexibility in the hindfoot. Specifically, is the subtalar joint completely correctable into a neutral position with or without supination of the forefoot? If such a reduction is possible, can it be achieved without associated significant forefoot supination? The overall approach to correction of deformity is therefore based on the flexibility of the foot: the presence of rupture of the PTT, the spring ligament, or the deltoid ligament, and the presence of any arthritis or secondary deformity of the midfoot.
Tenosynovitis
A common error in the management of the flexible flatfoot is failure to recognize the presence of tenosynovitis in patients who have a seronegative spondyloarthropathy. This condition is not necessarily associated with a flatfoot deformity, but if it is not treated with a tenosynovectomy, the tendon will be progressively infiltrated until rupture and deformity occur. In the more common presentation of tenosynovitis in the older patient, they present with pain along the course of the tendon generally associated with a mild flatfoot deformity. One has to decide whether or not an osteotomy of the calcaneus should be performed with the tenosynovectomy, and while this was not our original recommendation, one has to consider tenosynovitis as “prerupture” and an osteotomy should be performed in all cases. The tenosynovectomy does not need to be performed as an open procedure, but can be done endoscopically.
How to Manage the Failed Tendon Transfer
It is important to establish why the FDL tendon transfer procedure failed. Was the selected tendon transfer not strong enough for the specific foot deformity? Is the patient obese? Did preoperative assessment fail to detect some stiffness of the hindfoot? Is there additional deformity of the medial column? Should an osteotomy of the calcaneus have been performed, and if so, with lengthening or medial translation?
Fig. 15.1 presents an example of failure of a previous FDL tendon transfer. The patient was a 54-year-old active woman weighing 76 kg. The hindfoot was quite flexible, and with the heel reduced to neutral there was a fixed forefoot supination deformity of 20 degrees. No active inversion was noted, and pain was present along the medial border of the foot and ankle, in addition to sinus tarsi pain, the result of subfibular impingement. Traditionally, if the FDL has already been used in a tendon transfer, little is left to function as an invertor. The lateral side of the foot can be weakened with a peroneus brevis-to-longus transfer, which would have the added benefit of plantarflexing the first ray. The lateral radiograph (see Fig. 15.1A ) shows a break in the foot at both the TN and NC joints. Certainly a fusion of the NC joint could be added to the procedure without loss of too much function, but this procedure would not address the weakness and lateral foot pain. The patient had already undergone a successful triple arthrodesis on the opposite foot for management of a very rigid flatfoot deformity and wanted to avoid another arthrodesis. In this patient, uncovering of the TN joint was not so extensive as to necessitate a lengthening of the lateral column, whether by osteotomy or by arthrodesis. We do not like to use a subtalar implant in the adult, and although this would be a reasonable indication, this procedure still does not address the muscle imbalance. We therefore selected the anterior tibial tendon (ATT) for transfer, modifying the original Young procedure.
For this modified procedure, the navicular is exposed, and a thick osteoperiosteal flap is raised from the center of the navicular and retracted inferiorly (see Fig. 15.1C ). This flap can include the remnant of the PTT as well. A ridge with an overhang of approximately 5 mm is then prepared under the navicular with a curette or chisel, to function as a ledge for the transferred ATT. The retinaculum of the ATT is opened and released proximal to the ankle joint under direct vision. The ATT is then pulled below the navicular into the prepared slot (see Fig. 15.1D ) and sutured either using the osteoperiosteal flap or by adding a suture anchor into the body of the navicular. If an anchor is used, fluoroscopic guidance is recommended to ensure that it is in the navicular and not protruding into the joint. The remnant of the PTT can now be cut, sacrificed, or repaired, depending on the status of the tendon (see Fig. 15.1E ). In this patient, after these procedures there was slightly more forefoot supination than we would accept, and plantar flexion of the medial column was obtained with an opening wedge osteotomy of the medial cuneiform. The postoperative radiographs showed excellent restoration of the height of the arch of the foot, as well as coverage of the TN joint (see Fig. 15.1F and G ). We would say that the only downside to this procedure is very slight weakness of dorsiflexion. The insertion point of the ATT effectively moves more proximally, and the ATT functions as an invertor as well as a dorsiflexor of the foot. The second potential for complications with this procedure is a rupture of the ATT under the navicular, but this will only occur with inadequate preparation of the trough leaving rough edges behind which cause attrition of the tendon.
A similar example of the same problem is shown in Fig. 15.2 . The patient was a 61-year-old man who had been treated with a previous calcaneus osteotomy and a transfer of the flexor hallucis longus. The hindfoot was quite flexible, and the heel was in moderate valgus. On observation of the heel from posterior, the calcaneus osteotomy did not appear to have accomplished much. The elements of surgical decision making were similar to those in the case shown in Fig. 15.1 , in which an arthrodesis was an option; however, maintaining flexibility seemed preferable in that very active patient. The ATT was transferred under the navicular, and the calcaneus osteotomy was repeated. It was not necessary to include an osteotomy of the medial cuneiform. An important point is that the ATT is not detached from its insertion and is levered inferiorly under a prepared osteoperiosteal flap in the edge of the navicular. The extent of bone resected in this process can be appreciated on the anteroposterior radiographic view (see Fig. 15.2C ), in which the medial pole of the navicular has been removed to create the ridge under which to attach the ATT. It is important to leave no sharp edges on the ridge on the proximal navicular, because consequent damage to the ATT can lead to its rupture—a complication that has occurred in one of our own patients. When the rupture occurred, the patient noticed a sudden change in the foot at 8 weeks after surgery, and the ATT was no longer palpable dorsomedially. The rupture was repaired by reattaching the ATT to the dorsal central aspect of the navicular using a suture anchor.
How to Manage the Failed Bony Correction
Although errors in decision making and technique are inevitable, some of these are entirely avoidable. In the case illustrated in Fig. 15.3 , a 47-year-old woman was treated with a transfer of the FDL for correction of a flatfoot deformity. At 9 months after the surgery, she presented with considerable pain in the anterior ankle, and on an anteroposterior radiograph of the foot, marked loss of the TN joint space was readily apparent. The diagnosis of rheumatoid arthritis was confirmed with a positive rheumatoid factor titer, and a TN arthrodesis was used for correction.
A simple error in technique is illustrated in Fig. 15.4 . In the case depicted, an opening wedge osteotomy of the cuneiform was performed in conjunction with a transfer of the FDL tendon and a calcaneus osteotomy. The incorrect insertion of the bone wedge into the cuneiform, which fractured into the first metatarsocuneiform (MC) joint (see Fig. 15.4 ), is evident. This technical error can be avoided by inserting a guide pin in the cuneiform from dorsal to plantar and obtaining a lateral view radiograph to ensure the correct orientation of the osteotomy. The cuneiform inclines obliquely, and the osteotomy is not perpendicular to the axis of the foot but must be oriented to the axis of the cuneiform. The osteotomy cut must also be completed with the saw and not end in the cuneiform, because use of an osteotome to complete the cut may result in fracture extending out into the MC joint, which probably happened in this case. Although arthritis did not develop in this patient over 3 years of follow-up, the potential for this complication nonetheless remains worrisome. The “hump” evident in the midfoot was the result of the incorrect axis of the osteotomy, which unfortunately caused abutment and symptoms in shoes.
What is the optimal approach to treatment of the extremely flexible flatfoot? Is it always possible to avoid arthrodesis? In Fig. 15.5 , illustrating the case of a 44-year-old man with a unilateral flexible flatfoot, significant uncovering of the TN joint is apparent. The lateral radiographic view showed quite a lot of sag at the TN joint as well, with a fairly marked alteration in the talocalcaneal angle. In such cases, we have found that despite addition of a lateral column–lengthening procedure, instability may persist at the TN joint, increasing the likelihood of subsequent failure. Accordingly, although it is therefore always preferable to perform a tendon transfer with osteotomies, a TN arthrodesis is a good, reliable procedure and was performed in this patient with a successful result. The postoperative improvement in the talocalcaneal angle can be readily appreciated on the lateral view, even on a non–weight bearing radiograph (see Fig. 15.5D ). We suspect that with increased knowledge and skill with reconstruction of the spring ligament, that better support of the TN joint can be expected in conjunction with a lateral column lengthening procedure. The other option for treating the very flexible foot is a subtalar arthrodesis in conjunction with a transfer of the FDL and a Cotton osteotomy if required. When performing the subtalar arthrodesis, the joint is very slightly internally rotated to improve the arch of the foot. This may have the effect of adding to the forefoot supination, but this should be correctable with either a Cotton osteotomy or an arthrodesis of the first TMT joint.
Perhaps a similar result could be achieved with a lengthening arthrodesis through the calcaneocuboid (CC) joint, but this too has not been our experience, and persistent abduction of the forefoot may be present ( Fig. 15.6 ). An arthrodesis of the CC joint preserves slightly more medial column motion, and although it still locks up inversion and eversion, it is nevertheless preferable to the TN arthrodesis under similar circumstances. It is, however, associated with a much higher risk for nonunion of the CC joint. The case illustrated in Fig. 15.6 involved a number of errors, including perhaps that related to decision making—the extensive correction required was too much to expect from this procedure. Given the reliability of a lateral column lengthening through the neck of the calcaneus, we prefer this technique over a lengthening CC fusion given the higher nonunion rate and difficulty of revision. As with all procedures, however, care must be taken during execution of a lateral column lengthening to avoid complications, such as violation of the subtalar joint or CC joint, and subluxation of the cuboid. Persistent abduction of the foot was present, insufficient lengthening was achieved, and a nonunion of the joint further complicated the outcome, which was resolved with a triple arthrodesis.
Another similar dilemma is illustrated in Fig. 15.7 . In this case, the patient was a 57-year-old man with bilateral flexible flatfoot. Indeed, because of the extreme flexibility, the use of osteotomies to perform the correction was appealing. Nevertheless, perhaps because of just a sense of when things are more likely to go wrong, derived from experience, we selected a triple arthrodesis for correction. In realigning the foot intraoperatively, it was apparent that correction of the TN coverage was possible only if a lengthening of the lateral column was simultaneously performed. The triple arthrodesis was therefore accomplished with a bone block graft in the CC joint (see Fig. 15.7 ).
