Medial Soft-Tissue Release for Lateralising Calcaneal Osteotomy: A Cadaveric Study

Abstract

Background

Lateralising calcaneal osteotomy for pes cavus is generally regarded to be harder to shift than a medialising calcaneal osteotomy for pes planus. The aim of our study was to determine the structures which restrain a lateral shift.

Methods

Lateralising calcaneal osteotomy was performed on four soft-embalmed cadavers via a standard lateral approach and the lateral calcaneal shift was measured before and after the release of flexor retinaculum. Further exploratory dissection around the osteotomy site revealed the abductor hallucis muscle to be the main restraint to the lateral shift of the calcaneus. Subsequently, lateralising calcaneal osteotomy was performed on another four cadavers and the abductor hallucis muscle fascia as well as the plantar fascia was released. The lateral shift was measured before and after the fascia release, and compared with the results achieved following the flexor retinaculum release in the first four cadavers.

Results

Lateralising calcaneal osteotomy alone resulted in an average of 4.5-mm lateral shift in the first four cadaveric specimens. Releasing the flexor retinaculum led to a further 3-mm increase of lateral shift on average. In the next four cadaveric specimens, lateralising calcaneal osteotomy alone resulted in an average of 5.5-mm lateral shift. Release of abductor hallucis muscle fascia and the plantar fascia in these four specimens increased the lateral shift by an additional 7 mm on average. Hence, release of abductor hallucis muscle fascia resulted in an extra 4-mm shift on average compared with what is achieved with flexor retinaculum release.

Conclusions

Abductor hallucis muscle fascia was discovered to be one of the main structures limiting the lateral shift in lateralising calcaneal osteotomy. Release of fascia over this muscle as well as the plantar fascia should help in improving lateral shift. Further experimental and clinical research is necessary to confirm the findings of this pilot study.

Background

Lateralising calcaneal osteotomies (LCO) are commonly used for correction of pes cavus or subtle cavus deformities, either alone or in conjunction with other procedures [1, 2]. It is widely recognized that LCO is more difficult to shift than a medialising calcaneal osteotomy (MCO) for pes planus. As a result, multiple variations of the LCO are described in the literature to overcome the problem of under-correction often faced in the surgical correction of hindfoot varus [3]. This could be due to tight tissues/neuromuscular constraints that are seen in pes cavus in general [4, 5]. The exact identification of these limiting structures, however, lacks confirmation.

Bruce et al. concluded that lateralising, but not medialising, calcaneal osteotomies produced a significant reduction in tarsal tunnel volume [6]. This also highlighted the post-operative risk of tarsal tunnel syndrome, which is in agreement with prior case reports in literature. It was found that with a greater lateral shift measurement, the tarsal tunnel volume decreased, respectively. Krause et al. reported cases of tibial nerve palsy in two patients with Charcot–Marie–Tooth disease (one of the most common neurological conditions associated with pes cavus) [7]. It was suggested that the mechanism of the decrease in tarsal tunnel volume is due to the flexor retinaculum being tightened during the lateral shift. Walls et al. also reported cases of postoperative tarsal tunnel syndrome in pes cavus patients who received a lateralising calcaneal osteotomy and suggested that a prophylactic release of the tarsal tunnel would be beneficial [8]. VanValkenburg et al. reviewed 80 osteotomies and confirmed incidence of neurological deficit to be 34%, with a majority (59%) resolving fully by 3 months [9]. Interestingly, VanValkenburg et al. found no protective effect of tarsal tunnel release against neurological injury [9].

Based on the knowledge from Krause et al. [7] that the flexor retinaculum is tightened with a lateral shift, and evidence from Bruce et al. [6] that the tarsal tunnel volume is decreased with an increased lateral shift, it was thought that the tarsal tunnel (the flexor retinaculum included) was the main restricting force in achieving a good lateral shift in lateralising calcaneal osteotomies. Hence we hypothesized that releasing the tarsal tunnel prophylactically could allow both greater correction and decreased risk of post-operative tarsal tunnel syndrome. However, taking into the consideration the findings of VanValkenburg et al. [9] where tarsal tunnel release did not appear to be protective, we remained open to explore other potential tight structures around the location of the osteotomy. The purpose of this study addressed this, with the aim being to identify which anatomical structures restrain the shift in lateralising calcaneal osteotomies for pes cavus.

Methods

Approval was obtained to perform a cadaveric study on eight soft-embalmed below-knee cadavers (four pairs). Eight cadavers were divided into two groups of four specimens each. For the Group 1, standard osteotomy along with tarsal tunnel release was performed.

Calcaneal osteotomies were carried out by a single consultant orthopaedic surgeon using standard lateral approach. The incision was approximately 45° angle to the foot over the lateral aspect of the calcaneus, posterior to the posterior facet of the subtalar joint and anterior to the insertion of the Achilles tendon. Subcutaneous tissues and superficial fascia and veins were then dissected and retracted using a blunt Weitlaner retractor. This allowed adequate exposure of underlying structures. Deep dissection was continued until the visibility of the periosteal surface of the inferior lateral calcaneal wall was clear, retracting the peroneal tendons and the belly of the extensor digitorum muscle when appropriate.

The osteotomy itself was carried out as a translational type osteotomy, beginning with a periosteal incision parallel to the initial skin incision. The osteotomy was performed in an oblique orientation, by first starting supero-lateral and progressing from lateral to medial side while aiming towards the infero-medial cortex. The osteotomy of the near cortex was performed along the line of the skin incision. A lamina spreader was used to distract the osteotomy and divide the medial periosteum to complete the osteotomy. Following this, lateral displacement of the osteotomy was performed manually and fixed with two thick K-wires (2 mm each). The lateral shift was measured from the lateral displacement of the posterior fragment of the calcaneus to the lateral wall of the anterior fragment in millimetres using a tape ruler.

A tarsal tunnel release was then performed from the medial aspect of the foot on the first four specimens, with the hypothesis that this would allow space for a greater lateral displacement. This was done by making a medial incision, approximately 6 cm proximal to the medial malleolus, curving inferiorly and anteriorly to 2 cm posterior to the tibia and medial malleolus in standard fashion. Subcutaneous tissue was dissected and skin flaps were retracted in a similar manner to the previous lateral dissection until the flexor retinaculum was visible. The flexor retinaculum was then released and retracted. With careful dissection to the underlying soft tissue, the deeper structures including the tibial nerve and its branches (medial plantar nerve, lateral plantar nerve and the medial cutaneous nerves) were identified. The posterior tibial artery and vein were also identified and carefully released from their respective fascia and surrounding soft tissues.

On achieving the tarsal tunnel release, the osteotomy fragment was again displaced, and fixed with two thick K-wires (2 mm each) and the lateral shift re-measured. Results were continuously recorded by two observers.

Although it was expected that the tarsal tunnel release would improve lateral shift, when this was implemented it was not found to be the case. During interim analysis after four tarsal tunnel releases, it was inferred that there was no significant increase in the lateral shift. This contradicted our original hypothesis that tarsal tunnel (including the flexor retinaculum) is the main restricting structure. Having our hypothesis therefore challenged, it was then decided to explore other anatomical structures on medial aspect of foot in order to find the limiting restraint. This was achieved by extending the medial incision used for the tarsal tunnel release on the first four cadavers anteriorly and posteriorly, dissecting the underlying subcutaneous tissue and retracting the skin and soft tissue flaps accordingly (see Fig. 1a). By manually shifting the posterior part of the calcaneus laterally and simultaneously examining which structures were tightened on the medial side, it became clear that the main restraints for a good lateral shift were the abductor hallucis muscle and the plantar fascia, more so the former. These structures were found to be under great tension where the osteotomy was shifted laterally.

Fig. 1.

Exploration and release of the abductor hallucis muscle fascia and plantar fascia. a Medial side of a specimen in Group 1 showing method of exploration of the anatomical structures. A vertical incision is made through the abductor hallucis muscle fascia, b Group 2 specimen showing the medial skin incision and fascia incision for abductor hallucis muscle fascia and plantar fascia release after lateralising calcaneal osteotomy in a specimen in Group 2

When this was made clear, the first four cadaveric specimens that received the tarsal tunnel release became Group 1, with the next four specimens that would be dissected becoming Group 2. In Group 2, the lateralising calcaneal osteotomy was done in an identical way to the first group of specimens with a control measurement taken from the initial lateral shift; however this time, the abductor hallucis muscle fascia and plantar fascia were released following osteotomy, rather than the tarsal tunnel releases in Group 1. Although the abductor hallucis muscle itself was a very stretched and tightened structure, it would be insensible to divide the muscle belly when considering the potential patient outcomes. As the fascia of this muscle is thick, it was assumed that a difference in shift could still be achieved by releasing this alone.

To release the abductor hallucis muscle fascia, a separate longitudinal incision was made approximately 3–4 cm in length over the medial aspect of the calcaneus at the medial area of the plantar fascia (see Fig. 1b). This incision was similar to the one commonly used for performing a Steindler’s release. The underlying thick fascia over the abductor hallucis muscle was identified and carefully released using a scalpel and dissecting scissors, perpendicular to the muscle belly. The release was extended inferiorly to include the medial band of the plantar fascia. The lateral shift of the posterior calcaneus fragment was performed again and fixed with two K-wires as before. The shift was again measured in millimetres. This was done on all four specimens in Group 2 and the results were recorded.

The differences in amount of lateral shift were compared between two groups, using non-parametric Mann–Whitney U test.

Results

Lateralising calcaneal osteotomy alone resulted in an average of 4.5 mm (SD 0.577, range 4–5 mm) and 5.5 mm (SD 0.577, range 5–6 mm) lateral shift in the first four (Group 1) and last four (Group 2) cadaveric specimens, respectively. Subsequent release of flexor retinaculum in the first four cadavers (Group 1) added an average of additional 3-mm shift to produce a total shift of 7.5 mm (SD 1.732, range 5–8 mm) on average (see Table 1), which was not a considerable amount of shift to the calcaneus (see Fig. 2a).

Table 1.

Lateral calcaneal bone shift before and after flexor retinaculum release

	LCO alone (mm)	LCO + flexor retinaculum release (mm)	Difference (mm)
1	5	8	3
2	5	9	4
3	4	8	4
4	4	5	1
Average	4.5	7.5	3

LCO lateralising calcaneal osteotomy

Fig. 2.

Lateral shift in specimens from Group 1 and Group 2 after their respective anatomical releases. a Specimen from Group 1 showing a 6-mm lateral shift, b specimen from Group 2 showing a 12-mm lateral shift

However, after interim analysis and exploration which led to change in dissection method, release of the abductor hallucis fascia and the plantar fascia in the last four cadavers (Group 2) added an average of additional 7-mm shift to create a total shift of 12.5 mm (SD 0.577, range 10–14 mm) on average (see Table 2), which was a substantial amount of shift to the calcaneus (see Fig. 2b).

Table 2.

Lateral calcaneal shift before and after abductor hallucis and plantar fascia release

	LCO alone (mm)	LCO + abductor hallucis and plantar fascia release (mm)	Difference (mm)
1	5	12	7
2	6	14	8
3	6	14	8
4	5	10	5
Average	5.5	12.5	7

LCO lateralising calcaneal osteotomy

With the control average of 5.5 mm, the comparative adjusted average lateral shift in Group 2 after fascia release was 10.5 mm, almost double the initial measurement (Group 2—before vs. after release). Releasing the fascia over abductor hallucis muscle together with the plantar fascia increased the lateral calcaneal displacement by a further 4 mm on average in comparison with flexor retinaculum release (Group 2 vs Group 1). The difference in amount of lateral calcaneal shift in second group was found to be significantly greater than those of first group (p < 0.05). (Mann–Whitney U test).

A major factor which was difficult to standardize and could potentially have affected results was the force by which the posterior fragment of the calcaneus was shifted. All shifts were done manually by the surgeon meaning it was difficult to measure accurately the force used and whether or not it differed in each specimen. This left a potential margin for error and perhaps accounts for the 1-mm difference in average control measurements between Group 1 and Group 2.

Discussion

Although Bruce et al. suggest prophylactic tarsal tunnel release in patients undergoing a lateralising calcaneal osteotomy, and Krause et al. discussed the tightening of the flexor retinaculum; neither of these studies mentions the possibility of the tarsal tunnel being the cause of a restricted lateral shift [6, 7]. In fact, there is scarce literature to suggest the release of any particular structure to maximize the lateral shift, despite it being so well known that lateral shifts are more difficult to achieve.

In this study we attempted to identify structures which may restrict the lateral movement of the calcaneal bone during LCO procedure through examining a new method of soft tissue release, which we discovered during an exploratory soft tissue dissection around the osteotomy site, and comparing that with another method of soft tissue release recommended in LCO (i.e. tarsal tunnel release), using cadaveric specimens. Lateralising calcaneal osteotomy alone created a minimal 4.5–5.5-mm lateral shift to the calcaneal bone, and subsequent release of the flexor retinaculum provided a further 3-mm lateral displacement to the bone on average (a total of 7.5-mm shift on average). However, release of the abductor hallucis fascia in conjunction with plantar fascia added a considerable additional 7-mm lateral shift to the calcaneus to create a total of 12.5-mm shift on average. According to Jung et al., that magnitude of lateral shift can help to correct even severe cases of pes cavus deformities, as mild, moderate, and severe degrees of hindfoot varus are defined by tibio-calcaneal axis angle (TCA) of below 10°, 10°–20°, and above 20°, respectively, and their correction requires a lateral calcaneal displacement of approximately 5 mm to 10 mm, 10 mm, and 10–15 mm, respectively [10].

The results of our study suggest that abductor hallucis muscle is one of the main structures limiting the lateral shift of the calcaneus in LCO, and we believe that release of the fascia over this muscle as well as the plantar fascia should be an indispensable part of the corrective LCO for pes cavus to achieve optimal lateral shift.

When considering implementation of the method suggested in this investigation, it is vital to anticipate any potential complications or detrimental effects that could occur. One critical point which should be firstly noted is the fact that the increased lateral shift was achieved without releasing the tarsal tunnel. A question arises over the risk of postoperative tarsal tunnel syndrome, given that our suggested method should increase the lateral shift by an even greater extent than Bruce et al. achieved [6]. If this is the case, it may be a necessity to consider a prophylactic tarsal tunnel release in addition to the abductor hallucis muscle fascia and plantar fascia release. This could be done through the same medial incision. Although it may lead to a longer procedure for surgeons, it would allow a decreased postoperative risk of tarsal tunnel syndrome, while also allowing an increased lateral shift, therefore a potentially better outcome for patients. However, it also needs to be highlighted that VanValkenburg et al., while reviewing 80 cases of lateralising osteotomies with 34% incidence of neurological deficit, have not been able to demonstrate any protective effect of tarsal tunnel release [9]. Stødle et al., in their review of 18 osteotomies, were unable to link a greater reduction in tarsal tunnel volume with neurological deficit, and questioned whether saw blade, osteotome and traction are responsible for nerve injury [11]. Therefore, further research on how to avoid neurological injury while achieving optimal lateralising shift is welcome.

Furthermore, it was not possible to accurately tell whether the abductor hallucis fascia release would affect the function of the great toe. Although plantar fascia releases are commonly performed for patients with plantar fasciitis, abductor hallucis muscle fascia releases are rarely implemented [12]. However, in young patients with plantar fasciitis and associated neurological symptoms, abductor hallucis fascia release is a textbook recommendation with recovery to full activity 3 months after release of both plantar fascia and abductor hallucis fascia [13]. Labib et al. also released the abductor hallucis muscle fascia in patients with heel pain triad, and function of the great toe was not affected in any patients on follow-up [14].

There are few limitations of this study; first, is the small sample size of the study. We have used non-parametric statistics to address the non-normal distribution of data associated with small sample sizes. Second, the dissection was performed on cadaver specimens, and there is a need to confirm the results in clinical studies. Patients with pes cavus may provide slightly different results due to factors such as the height of their arch, the stiffness of structures and the ability to shift calcaneus. Hence, authors acknowledge that clinical studies involving cavovarus patients will be necessary to confirm these findings. Nevertheless, even in these cadaveric specimens, a lateral shift of the osteotomy was still more difficult than a medial shift, thus it did reflect normal clinical practice.

Additionally, lack of radiographic imaging led to a much less accurate method of measurement, using a tape ruler. Although the margin of error was only between 0 and 1 mm, accuracy could still be improved using CT, MRI or X-ray, which would be available in a clinical setting. This would also avoid any discrepancies in bending the tape measure or the orientation of the measurement, as well as being more numerically specific. Therein lies the need for continued research in this field, as calcaneal osteotomy will foreseeably remain an important reconstructive option in the correction of cavovarus foot deformity. Our study can be considered as a good pilot research, although the authors understand that a well-designed serial sectioning study with larger specimen sample size, and more accurate measurement methods will be necessary to confirm the findings and support clinical research on this area.

Conclusion

Within the limits of a pilot research, our study demonstrated that the lateral shift improved significantly after medial soft-tissue releases in LCO. We conclude that releasing the fascia over abductor hallucis and plantar fascia should be recommended when performing a lateralising calcaneal osteotomy if further experimental and clinical studies confirm our findings.

Author contributions

Concepts: KKD, RB, IS, KS, Design: KKD, RB, KS, Definition of intellectual content: KKD, RB, KS, Literature search: KKD, RB, IS, KS, Experimental studies: RB, KS, Data acquisition: RB, KS, Data analysis: KKD, RB, IS, KS, Statistical analysis: KKD, Manuscript preparation: KKD, RB, IS, Manuscript editing: KKD, IS, KS, Manuscript review: KKD, RB, IS, KS, Guarantor: KS.

Compliance with Ethical Standards

Conflict of interest

The authors declare that they have no conflict of interest.