1

The tarsometatarsal (TMT) joint complex connects the midfoot to the forefoot and consists of joints formed by the bases of the metatarsals with the cuneiform and cuboid bones. The lateral portion of TMT joint complex exhibits relatively greater mobility and functions as a shock absorber, whereas the medial and middle columns are comparatively rigid . The rigidity of medial and middle columns is critical for normal foot biomechanics and for maintaining transverse arch stability .

Since the early 1900s, medial column hypermobility has been implicated in the pathogenesis of hallux valgus, leading to the widespread adoption of first tarsometatarsal (TMT) arthrodesis following the work of Lapidus ^, . To improve construct stability and reduce complications, extension of the fusion to the second metatarsal and intercuneiform joints has been advocated ^,,, , although its necessity remains controversial.

Similarly, medial column TMT arthrodesis is increasingly used in the treatment of Lisfranc injuries, where primary arthrodesis has been shown to provide clinical outcomes and return-to-sport rates comparable or superior to internal fixation ^,, . Despite its frequent use, the biomechanical consequences of extending medial column fusion to adjacent joints remain incompletely understood.

Adjacent joint arthritis is a well-recognized complication of arthrodesis, particularly involving the tibiotalar and subtalar joints ^,,,, . This has been attributed to altered load transfer and increased mechanical stress in adjacent joints following fusion . However, biomechanical data specifically addressing adjacent joint loading after tarsometatarsal arthrodesis remain limited ^,,, . In addition to changes in joint loading, alterations in plantar pressure distribution represent an important and clinically relevant manifestation of load redistribution following arthrodesis.

Because direct assessment of joint loading after arthrodesis is limited in clinical settings, finite element analysis has been used to investigate the biomechanical effects of tarsometatarsal fusion. Existing finite element analysis studies have mainly focused on fixation constructs ^,, or combined first and second TMT arthrodesis , providing limited insight into how selective medial and middle column fusions affect adjacent joints and plantar pressure. Therefore, this study aimed to evaluate changes in joint reaction forces, plantar pressure distribution, and tibiotalar contact pressure across different phases of gait following selective TMT arthrodesis configurations.

2

2.3

Midfoot arthrodesis models ( Fig. 1 )

In the finite element analyses, the intact model was evaluated first and served as the reference condition. Subsequently, six arthrodesis configurations (Models I- VI) were generated to represent commonly used clinical fusion strategies. Arthrodesis was simulated by removing the articular cartilage at the corresponding joints, assigning bone material properties to the joint interfaces, and applying bonded contact conditions between the fused surfaces. This approach eliminated relative motion at the fused joints while preserving load transfer across the fusion sites. No fixation hardware was modeled in order to isolate the biomechanical effects of joint fusion itself.

In Model I, fusion was applied solely to the first tarsometatarsal (TMT) joint. Model II represented isolated fusion of the medial cuneiform- second metatarsal joint. Model III combined these two fusions. In Model IV, fusion of the second TMT joint was added to Model III. Model V further included fusion of the medial- intermediate intercuneiform joint. In the final configuration (Model VI), fusion of the third TMT joint and the intermediate- lateral intercuneiform joint was added, resulting in the most extensive fusion construct ( Fig. 1 ).

Midfoot arthrodesis models. Model I: First TMT joint. Model II: Medial cuneiform- second metatarsal joint. Model III: Combination of Model I and Model II. Model IV: Addition of second TMT joint to model III. Model V: Addition of medial-intermediate cuneiform joint to model IV. Model VI: Addition of third TMT joint to Model V.

3

The intact model served as the reference condition for all biomechanical comparisons.

3.1

Joint reaction force alterations ( Fig. 2 )

Across all gait phases, joint reaction force (JRF) alterations were predominantly observed within the midfoot in all models. The medial- intermediate intercuneiform and calcaneocuboid joints demonstrated the most pronounced increases in JRF. Forces at the tibiotalar and subtalar joints remained largely unchanged (generally <5% deviation from the intact condition).

Isolated medial cuneiform- second metatarsal fusion (Model II) consistently produced the smallest deviations from the intact condition, with JRF changes remaining below 10% across all evaluated joints and gait phases. In contrast, inclusion of the third tarsometatarsal (TMT) joint (Model VI) resulted in the greatest alterations, with increases in JRF reaching approximately 25–32% at the calcaneocuboid and lateral column joints, particularly during late stance. Intermediate configurations (Models III- V) demonstrated comparable JRF patterns, with differences between these models generally remaining below 5% ( Fig. 2 ).

Joint reaction force comparisons relative to the intact model. (a) Early stance. (b) Midstance. (c) Late stance.

3.2

Plantar pressure distribution changes ( Fig. 3 )

Relative to the intact model, plantar pressure increases were predominantly observed in the first and second metatarsal regions across most arthrodesis configurations. These changes were phase dependent and most pronounced during late stance.

Model II showed minimal deviation from the intact plantar pressure pattern, with peak pressure changes generally remaining below 10–15 kPa across all gait phases. In contrast, Model VI demonstrated the largest plantar pressure alterations, reaching approximately 25–30 kPa during midstance and up to 50–60 kPa during late stance. Models III, IV, and V exhibited highly similar plantar pressure distributions, with inter-model differences remaining small compared with those observed in Model VI ( Fig. 3 ).

Plantar pressure distribution comparisons relative to the intact model. (a) Early stance. (b) Midstance. (c) Late stance.

3.3

Tibiotalar contact pressure redistribution ( Fig. 4 )

Total tibiotalar joint reaction forces did not change substantially following tarsometatarsal arthrodesis; however, a consistent redistribution of tibiotalar contact pressure was observed across all arthrodesis models. Pressure alterations followed a characteristic anterolateral- posteromedial pattern on the talar articular surface throughout the stance phase.

The magnitude of tibiotalar pressure redistribution increased with the extent of fusion. Model II demonstrated the smallest changes, whereas Model VI exhibited the greatest alterations, with peak contact pressure differences reaching approximately 300–350 kPa during midstance and up to 500–550 kPa during late stance ( Fig. 4 ).

Tibiotalar contact pressure distribution comparisons relative to the intact model. (a) Early stance. (b) Midstance. (c) Late stance.

4

The most important finding of the present study is that the biomechanical consequences of tarsometatarsal arthrodesis strongly depend on the extent of fusion. While isolated medial cuneiform- second metatarsal fusion resulted in minimal alterations in joint reaction forces, plantar pressure, and tibiotalar contact mechanics, extension of fusion to the third tarsometatarsal joint produced pronounced changes, particularly involving the lateral column and adjacent joints. Among these, the calcaneocuboid joint consistently demonstrated increased joint reaction forces, highlighting its susceptibility to load redistribution following more extensive fusion constructs. Notably, although total tibiotalar joint reaction forces remained largely unchanged, a consistent redistribution of contact pressure along an anterolateral- posteromedial axis was observed, with increasing magnitude as the fusion construct was extended. These findings suggest that selective fusion strategies may substantially influence load transfer patterns beyond the fused joints.

TMT joint arthrodesis is commonly employed in the treatment of various forefoot and midfoot pathologies. The first TMT joint arthrodesis (Model I), popularized by Lapidus, is widely used in the surgical management of advanced hallux valgus . In selected cases, fusion may be extended to additional medial column articulations to enhance construct stability or address midfoot instability ^, . Isolated fixation between the medial cuneiform and the second metatarsal represents a treatment option, particularly for ligamentous Lisfranc injuries (Model II) . Furthermore, arthrodesis of the third TMT joint should be considered in degenerative conditions or following Lisfranc injuries (Model VI). Collectively, the arthrodesis configurations evaluated in this study span a range of fusion extents and were included to provide a structured biomechanical framework for assessing the biomechanical effects of incremental midfoot fusion on load transfer and adjacent joint mechanics.

Joint reaction force (JRF) is the vectoral sum of all the forces acting on a joint, and increase in JRF can be considered as a key factor in osteoarthritis pathogenesis . Although there are no established cut-off points for alterations in JRF in terms of clinical relevance, the investigation of changes in this study implicates several findings. One of the main findings is that without fusion of medial-intermediate intercuneiform joint, there was a constant increase of JRF on this joint. Intercuneiform joint issues, especially instability, was emphasized in hallux valgus ^, , and adding a fusion of this joint to the construct was recommended to avoid potential complications related to this joint . The present findings biomechanically support consideration of intercuneiform fusion in selected cases by demonstrating increased joint loading when this joint remains unfused.

The calcaneocuboid joint is a key component of the transverse tarsal joint complex and plays a critical role in maintaining the lateral longitudinal arch. Although isolated pathology of this joint is relatively uncommon, it has been previously reported that calcaneocuboid joint arthritis may progress to tibiotalar joint arthritis and contribute to chronic pain over the long term . In this context, the present findings demonstrate increased loading of the calcaneocuboid joint following more extensive TMT arthrodesis, suggesting a potential biomechanical mechanism by which adjacent joint degeneration may be promoted. Nevertheless, the clinical implications of this altered loading pattern require confirmation in future clinical studies.

The total JRF at the tibiotalar joint did not change substantially across any arthrodesis model. However, analysis of contact pressure distribution revealed a consistent shift in pressure toward an anterolateral- posteromedial direction in nearly all models. This redistribution was minimal in Model II (isolated medial cuneiform- second metatarsal fusion) and tended to increase with the number of fused joints, reaching its greatest magnitude in Model VI across all phases of gait. Previous biomechanical studies investigating ankle fractures or malalignment have demonstrated marked alterations in tibiotalar contact pressures ^,, . Although the magnitude of pressure change observed in the present study was quantitatively smaller than that reported in fracture models (maximum 558 kPa during late stance in Model VI), we believe that these alterations may nonetheless have clinical relevance. While no definitive threshold for clinically significant tibiotalar pressure changes has been established, even subtle modifications in cartilage load transfer may predispose to secondary pathological processes, including cartilage degeneration and the development of ankle arthritis.

We found that inclusion of the third TMT joint in the fusion construct, particularly affecting the lateral column joints (calcaneocuboid and cuboid- lateral cuneiform), resulted in increased JRF and exerted the greatest influence on both tibiotalar joint loading and plantar pressure distribution. Although incorporation of the third TMT joint may be necessary in certain clinical scenarios, particularly following Lisfranc injuries, it is not routinely included in fusion constructs except in cases of isolated pathology such as fracture or degenerative arthritis. In a cadaveric study, Kim et al. demonstrated that inclusion of the third TMT joint significantly altered the kinematics of adjacent joints and highlighted the importance of carefully evaluating the indication for extending the fusion to this joint . The findings of the present study are consistent with and support this recommendation.

We observed minimal changes in JRF at the first metatarsophalangeal joint following TMT arthrodesis. Previous studies have reported conflicting findings: while one study demonstrated a 25% increase in first metatarsophalangeal joint dorsiflexion following first TMT joint arthrodesis , Kim et al. did not observe a similar effect . When evaluating JRFs at the first metatarsophalangeal joint, our results are more consistent with those reported by Kim et al. Nevertheless, increased plantar pressure was observed in the first and second metatarsophalangeal joint regions in all models except Model II, particularly during the midstance and late-stance phases. These findings suggest that TMT arthrodesis may influence plantar pressure distribution even in the absence of substantial changes in joint reaction forces at the metatarsophalangeal level, and this potential effect should not be overlooked.

Analysis of plantar pressure revealed that the greatest increases occurred at the first and second metatarsophalangeal joint level, measuring approximately 20–30 kPa during midstance and 50–60 kPa during late stance. Previous studies have reported the minimal detectable change in plantar pressure in clinical settings to range between 30 and 80 kPa , while a threshold of approximately 200 kPa has been adopted in many studies as clinically relevant . In the present study, none of the observed pressure changes reached these established thresholds. Nevertheless, forefoot ulceration following midfoot arthrodesis has been previously reported . Therefore, awareness of potential alterations in plantar pressure remains particularly important, especially in high-risk patient populations such as individuals with diabetes mellitus.

The study is not without limitations. First, as the findings are derived from computational analyses, they are inherently based on modeling assumptions. Consequently, real-life clinical conditions may differ due to numerous patient- and case-specific factors. In addition, only a single CT-based model obtained from a healthy individual was used for the finite element analyses. Given that the clinical relevance of TMT arthrodesis is most pronounced in conditions such as hallux valgus and Lisfranc injuries, the use of models incorporating pathological anatomy might yield different biomechanical results. The use of a neutral, non-pathological foot model represents a simplification that facilitates comparative biomechanical analysis. Therefore, the findings should be interpreted as mechanistic insights rather than direct simulations of specific pathological conditions such as hallux valgus or Lisfranc injury. Bone tissue was modeled as a homogeneous structure with uniform material properties. While this approach facilitated simplification of the analyses, it should be noted that cortical and trabecular bone possess distinct mechanical characteristics, which were not differentiated in the present model. Similar simplifications were applied to the modeling of arthrodesis. Joint fusions were not simulated using fixation hardware; instead, the corresponding articular surfaces were fully bonded through contact definitions within the finite element framework. This represents a deliberate modeling choice and constitutes a limitation of the study, as the analysis was designed to investigate construct-specific biomechanical behavior following successful fusion rather than to simulate fixation stability or early postoperative mechanics. In clinical arthrodesis procedures performed with screws or plates, subtle differences in joint rigidity, degrees of freedom, and force transmission may occur. This modeling approach also precluded evaluation of the effects of different fixation materials on overall foot biomechanics. Finally, all fusion models were created without introducing any changes in alignment, preserving the native anatomical position. While this strategy allowed for direct comparison among different arthrodesis configurations, it does not fully reflect clinical practice, where corrective procedures that alter alignment are frequently performed in conjunction with fusion. Accordingly, the biomechanical consequences observed in this study should be interpreted within the context of these modeling constraints.