Impaired gait kinematics in type 1 Gaucher’s Disease

Abstract

Type 1 Gaucher’s disease (GD1) is traditionally regarded as “non-neurological”. Spatiotemporal and kinematic analysis of gait was carried on thirteen GD1 patients and thirteen healthy controls. We identified a previously unknown subclinical reduction of amplitude of movements in GD1. Articular excursion of ankle, knee and hip was reduced during the swing phase of gait (p <  0.0001). Furthermore, the excursion of the knee appeared also significantly more asymmetric in GD1 patients (p = 0.02). Correction for age, BMI and basal walking speed did not modify the significance. Accordingly to the recent observations that GD1 predisposes to Parkinson’s disease, the impaired and asymmetric gait kinematics that we observed might be interpreted as a form of extrapyramidal involvement.

INTRODUCTION

Gaucher’s disease is an autosomal recessive lysosomal storage disorder caused by a defective activity of the enzyme beta-glucocerebrosidase (GBA) [1]. It is the most common lysosomal disease affecting about 1/50, 000 in the general population. In type 1 Gaucher’s disease (GD1) nervous system involvement is not common, whereas it is always present in type 2 and 3 GD [2]. In the last few years, both biological and epidemiological evidence has arisen to show that in GD1 there could be some disruption of the normal neuronal functioning, causing a susceptibility to polyneuropathy [3], autonomic nervous system alteration [4] and to neurodegenerative diseases with synuclein accumulation such as Parkinson’s disease (PD) and Lewy Body Dementia [5–7]. Consistently, it has been reported that both GD1 and its heterozygous carrier state represent a risk factor for PD [5, 6].

Accordingly, McNeill et al. showed that GD1 patients with or without frank parkinsonism show subtle clinical signs of neuronal involvement such as cognitive impairment, hyposmia and higher UPDRS-III scores [5]. These findings support the hypothesis of neuronal degeneration and motor impairment in GD1. We hypothesize that a highly refined approach such as a Motion Analysis (MA) could help to define the possible motor impairment in GD1 patients, both by quantifying the differences in gait parameters and by obtaining insights on the kinematics of gait. To address this issue, we performed a MA study of the gait cycle in GD1 patients and healthy controls. For the first time we report the impairment of kinematic gait parameters in GD1 patients.

MATERIAL AND METHODS

Subjects and clinical assessment

This study was approved by the local Ethics Committee and informed consent was obtained from all participants according to the declaration of Helsinki. We screened 20 patients with GD1 followed at Federico II University Hospital. Seven patients were ruled out because of either musculoskeletal, neurological, or internal disorders (Table 1). Thirteen controls matched for age, sex, BMI and level of instruction were selected among the hospital staff. All the participants underwent a full internal, neurological and orthopedic examination. Signs of peripheral nervous system involvement were accurately ruled out, as well as joints, muscles and ligaments impairment. No haematologic signs were detected. The presence of hyposmia was anamnestically ruled out. A complete UPDRS-III score as well as Mini Mental State Examination (MMSE) was also carried out. Height and weight were measured and BMI was calculated. All the participants had bone MRI, BMD (Body Mineral Density) or both. All the participants underwent a MA assessment performed in the morning, in a single session that lasted approximately one hour.

Motion Analysis assessment

The marker-set included 42 retroreflective spherical markers placed directly onto the skin on standard body landmarks according to the Davis Modified protocol [8]. Markers were recorded with an eight-camera MA system (Qualisys, Sweden) at a sampling rate of 120 Hz. After adequate practice, all participants were instructed to walk at a self-selected speed for 10 meters. For each subject the six best gait cycles (out of aminimum of ten cycles) were selected and averaged. The kinematic gait data was retrieved from the averaged gait cycle. The gait cycle was defined as an interval from one heel contact to the next contact made by the same heel. Specific softwares (QTM^®; Visual 3D^®) were used to define the skeletal segments, to detect the gait events (required to identify two subsequent gait cycles) and to obtain a gait clinical report. The Range of Motion (RoM) (defined as the magnitude of rotatory motion that a body segment can move through) was obtained analyzing the joint excursion on the sagittal plane for the lower limb (thigh (T), knee (K), and ankle (A)) and calculating the minimum and maximum angular values, corresponding to the maximum flexion/extension peaks. Then, we calculated the Δ values as the difference between two consecutive peaks in the gait cycle, so as to sunder the total flexion/extension RoM into 4 sub-segments (TΔa-b-c-d, KΔa-b-c-d, AΔa-b-c-d) corresponding to four moments of the entire joint excursion (Fig. 1), normalized for the 100% of the gait cycle [9, 10]. Spatiotemporal gait variables are listed in Appendix 1. Since each of the four Δ of RoM has a specific biomechanics significance, it allows to better identify the angular values expressing the real variation of joint functionality.

A symmetry index was calculated for the Δ values [9] as follows: $i_a=\left|\frac{\frac{(\mathit{Left}X\Delta y-\mathit{Right}X\Delta y)}{(\mathit{Left}X\Delta y+\mathit{Right}X\Delta y)}}{2}\right|$

Statistical analysis

A statistical analysis software was used (SPSS Inc^®, USA). Results were expressed as means and range or 95% C.I. unless otherwise indicated. Comparisons between patients and controls were made using the Mann – Whitney U-test. Generalized linear regression (GLR) analyses were used to allow for possible confounders using gait measurements as dependent variable. Two-sided p values = 0.05 were considered statistically significant.

RESULTS

Thirteen healthy controls (five males, eight females), age, sex and body weight matched, and thirteen GD1 patients (five males and eight females) were enrolled (Table 1). The GD1 and control group were clinically comparable in their neurological and orthopedical condition. No statistically significant difference was found between GD1 and controls in MMSE and in all sub-items included in UPDRS-III. The analysis of the spatiotemporal parameters of gait failed to show any difference between the two groups (data non shown). On the other hand, the kinematic parameters as express by AΔc (p = 0.029), KΔc (p <  0.0001), KΔd (p <  0.0001) and TΔd (p <  0.0001) were statistically significantly lower in GD1 patients (Fig. 1). Very interestingly, KΔd was significantly more asymmetric in GD1 patients as compared to controls (i_aKΔd [GD1] = 0.0582 Vs i_aKΔd [Ctrl] = 0.0262; p = 0.02). In GLR models, after adjustment for age, BMI and basal walking speed statistical significances were maintained for AΔc (p <  0.041), KΔc (p <  0.001), KΔd (p <  0.001) and TΔd (p <  0.001). Moreover, KΔd remained significantly more asymmetric in GD1 patients as compared to controls (p = 0.009).

DISCUSSION

We have found evidence of altered gait pattern in GD1 patients as documented by subclinical kinematic changes. The parameters found to be statistically significant were all biomechanically and biologically consistent with diminished amplitude of the joint excursion of the leg during the swing and pre-swing phase of gait (subsegmental RoMs). More specifically, smaller AΔc and KΔc denoted a reduced extension of ankle and flexion of knee, respectively, during the early swing phase. Consistently, KΔd and TΔd indicated a smaller degree of extension of the knee and hip, respectively, during the late phases of the swing. In a nutshell, gait alterations were more evident in the pre-swing and swing phase, when the limbs are not touching the ground. In conclusion, although the spatiotemporal parameters were not different between GD1 patients and controls, the MA demonstrated an alteration of the gait kinematics, as shown by the reduced amplitude of the leg joints. Altered gait kinematics, along with not statistically significant differences of spatiotemporal parameters, might be compatible with a very early gait pathology with a functional compensation.

There is more than one possible interpretation of our findings. It may possible that in GD1 patients, a population with an increased risk to develop PD [5, 6], the diminished amplitude of the joint excursion of the leg might reflect an early preclinical involvement of the extrapyramidal system. Although the licterature on PD gait analysis is controversial, (possibly because of the heterogeneity of the patients and of the outcomes), a reduced angular excursion is found. Some studies showed this at the ankle joint [11], while others points toward prominent involvement of the hip [12] or of the knee [13]. Furthermore, studies in off patient [14] showed the alteration to be spread across the whole lower limb.

The fact that some kinematic parameters were found to be significantly more asymmetric in our patients as compared to controls might also be interpreted within the framework of an extrapyramidal involvement [12].

More intrestingly, Mirelman’s et al. [15], using an accelerometer system, compared the spatiotemporal parameters in asymptomatic carriers of LRRK2 G2019S mutation and healthy controls and found a statistical significant difference in stride variability (but not in stride time, stride length and gait speed) only in challenging condition (dual task and fast walk). They hypothesized that the compensatory mechanisms at work during normal walk (whereby the normal spatiotemporal parameters) became insufficient during more challenging conditions. Similarly, in our population of GD1 patients, a functional compensation could be hypothesized whereby the spatiotemporal parameters are normal while a more sensitive measure, such as the kinematic parameters, is compromised [16, 17]. Another possibility is that the gait abnormalities observed in GD1 patients may reflect a pathophysiological condition unrelated to extrapyramidal involvement. This issue can only be solved by prospective follow-up of these subjects. Finally, another possible explanation of observed gait abnormalities might be a bone involvement. However, normal clinical examination, no major alterations at MRI and/or BMD and the constant pattern of gait involvement make this hypothesis unlikely.

In conclusion, our observation is the first objective demonstration of gait impairment, possibly of neurological origin, in GD1. This has relevant implications on both the classification of the disease and the phenotype definition.

CONFLICTS OF INTEREST

AB has received honoraria from Genzyme and Shire and sponsorships for participation to scientific meetings by Actelion, Genzyme and Shire. Contributions for scientific research on rare lysosomal diseases were delivered by Genzyme and Shire to the Department of Clinical Medicine and Surgery, Federico II University of Naples. The other authors have no financial disclosures or conflicts of interest.