CASE REPORT
The patient was a 65-year-old Japanese man with diabetic neuropathy, who had undergone multiple toe amputation (second, third, and fourth phalanxes and at least part of the metatarsal) in left limb in May 2013 (when he was 60-year-old) and single toe amputation (third proximal phalanx) three year later, May 2016. The patient had comorbidities of critical limb ischemia and received hemodialysis. Although diabetic-related foot deformity, including Charcot foot, was not confirmed, left foot displayed clow toe after the surgery. Joint mobility in both forefoot and hindfoot was limited based on physical assessment by experienced medical doctor. He could walk independently on flat surfaces without using an ambulatory assistive device. Given the multiple toe amputations in the left foot, henceforth we have used the term “ipsilateral” for the left limb and “contralateral” for the right limb.
First, to determine whether the patient displayed a decline in walking capacity, we evaluated spatiotemporal gait parameters during level walking using a laser range sensor-based leg tracking system5 (Fig. 1A ). This new technology-based gait assessment has been applied to characterize ambulation in patients with neurodegenerative and musculoskeletal diseases6-8. As this analysis system evaluates gait parameters in the right and left limbs separately (Fig. 1B ), it provides insight into amputation-related gait alterations. As expected, this patient displayed poor walking capacity compared to the reference value derived from patients with diabetic neuropathy without amputation4, 9, 10, as evidenced by slower gait velocity, shorter stride length, and shorter step length, especially the ipsilateral step length (Fig. 1C ). These data indicate that patients who have undergone multiple toe amputation display gait disorders with inter-limb asymmetry in step length.
As partial foot amputation influences the center of pressure (COP) and GRF1, we determined whether toe amputation influences these kinematic changes. We used a validated custom-made sensor-mounted insole to assess the COP and GRF (Fig. 2A )11. This system can measure a force similar to the vertical GRF by summing the forces of 15 sensors attached to the insole. The system can also estimate the COP positions in the anteroposterior and mediolateral directions by calculating the weighted average of the sensor positions and forces. We found that the ipsilateral foot displayed reduced COP excursion, which was derived from a lack of push-off motion by the forefoot (Fig. 2B ). In addition, the COP during the terminal stance in the ipsilateral foot was localized at the first metatarsal, which was not observed in the COP of the contralateral foot and in healthy adults (Fig. 2B ). Of note, contralateral foot also displayed reduced COP excursion compared to healthy adult, suggesting that foot function in contralateral limb is not necessarily normal. The confirmed inter-limb difference in the COP was consistent with the GRF data showing inter-limb asymmetry (Fig. 2C ). On comparison with healthy adults, our patient displayed lower first and second peak values of GRF in the ipsilateral limb (Fig. 2C ). Notably, the first peak GRF in the contralateral limb was higher than that in healthy adults. These findings indicate that the lack of push-off in the ipsilateral limb is related to increased first peak GRF in the contralateral limb.
To further support the biomechanical inter-limb asymmetry, we recorded level walking using a 60-fps stationary camera (HDR-CX550V; Sony Corp, Sony Marketing Inc, Tokyo, Japan) and performed kinematic analysis based on the whole-body skeleton visualized using the OpenPose system12. We focused on the gait cycle from the ipsilateral terminal stance (i.e., contralateral heel contact), which corresponds to the observed alterations in the COP and GRF. This decision was made based on previous biomechanical evidence that the drop-off effect during the terminal stance in one limb leads to increased GRF in the contralateral limb3. Fig. 3 shows the results of the gait movie analysis in the sagittal plane. The hip extension angle during the terminal stance in the ipsilateral limb was lower than that in the contralateral limb (12.7° vs. 27.7°, as measured by Image J software).