Effects of Robot-Assisted Gait Therapy on Urodynamic Changes in the Subacute Phase After Spinal Cord Injury: A Prospective Study

Introduction

Multiple conditions can lead to the development of neurogenic bladder, including stroke, multiple sclerosis, cerebral palsy, hydrocephalus, and spinal cord injury (SCI) [1]. Approximately 70% to 84% of patients with spinal cord injury (SCI) have varying degrees of bladder impairment, the type of impairment depends on the level of damage, and its characteristic depends on the injury level [2]. Loss of supraspinal coordination and impaired reflexes in both the autonomic and somatic nervous systems following SCI can result in various bladder disorders, including involuntary reflex contractions of the detrusor during the filling phase (neurogenic detrusor overactivity) and detrusor contractions associated with concomitant external sphincter contraction (detrusor sphincter dyssynergia). Bladder function in incomplete SCI is variable and may resemble that in complete SCI, presenting with urinary urgency or preserved voluntary control [3]. Urodynamic studies remain the standard method for evaluating lower urinary tract dysfunction in this population.

Bladder management is one of the main goals in SCI rehabilitation. Inadequate management of neurogenic bladder can lead to incontinence, renal impairment, urinary tract infections, urolithiasis, and other complications, in addition to diminished quality of life and increased healthcare costs [4–6]. Optimizing management strategies is therefore essential to achieve favorable long-term clinical outcomes [3].

In the able-bodied population, physiotherapy interventions for urinary incontinence typically focus on strengthening the pelvic floor muscles, which also contribute to trunk stability while walking. However, the potential role of gait training in improving lower urinary tract function in individuals with SCI remains unclear [7].

Robot-assisted gait therapy (RAGT) is recently-developed therapeutic option for gait therapy in SCI patients [8] and other neurologic disorders [9,10]. RAGT can also prevent secondary problems associated with physical inactivity in SCI patients and can improve pain, spasticity, and cardiopulmonary, and urinary and bowel function, consequently enhancing patient functioning [11,12]. However, most studies assessing the effect of RAGT use on bladder function were based on self-reported data from patients with SCI or on responses to specific questionnaires to quantify changes in bladder function [13–16]. Only Wiliams et al assessed the effect of RAGT on lower urinary tract function using urodynamic examination in 5 of the 12 included SCI patients [17,18]. The small sample size of that study did not allow for definitive conclusions regarding the impact of RAGT on lower urinary tract function [17,18]. Our study is the first to be conducted in a large cohort of patients in the subacute phase after SCI, investigating the influence of locomotor training on lower urinary tract function, and the first to compare gait training using RAGT with dynamic parapodium training (DPT). DPT is a dynamic device that enables the patient to stand and simulate walking by swinging the trunk, and it is still used in rehabilitation in some countries.

The primary aim of this study was to evaluate the effect of RAGT on urodynamic parameters in patients with SCI. Secondary aims included assessing differences in urodynamics findings between patients with complete and incomplete SCI, examining the impact of different levels of SCI on urodynamics, and comparing gait training using RAGT versus DPT.

Material and Methods

STUDY PROTOCOL:

This was a single-center and single-blinded study. The study was conducted after receiving approval of the Bioethics Committee (Ethical Board of the District Medical Chamber in Szczecin, Poland, Nr OIL-Sz/MF/KB/452/05/07/2018; Nr OIL-Z/MF/KB/450/UKP/10/2018). The study was carried out in the Health Resort Kamień Pomorski S.A. Research Institute for Innovative Methods of Rehabilitation of Patients with Spinal Cord Injury in Kamień Pomorski.

STUDY POPULATIONS:

Patients were recruited through self-selection from all over the country. All patients who qualified for the study underwent earlier neurological rehabilitation in neurological rehabilitation departments. Participation in the study was voluntary. All participants provided written informed consent before enrolment.

The following inclusion criteria were applied:

Patients with SCI were excluded due to:

INTERVENTIONS:

The therapy program lasted 7 weeks and consisted of a 2-stage course: 3 weeks separated by a 7-day break. The program was conducted 6 days per week.

Participants were allocated into 2 groups: the control group (S0), which received conventional gait therapy using DPT, and the study group (S1), which received RAGT. A physiotherapist who was not involved in the treatment process (blinded investigator) was responsible for the group allocation process. Simple randomization was performed by tossing a coin. The assessor (a urologist) was blinded to group allocation. The S1 and S0 group underwent 30-min sessions of RAGT with exoskeleton EKSO-GT (model EKSO 1 by EKSO Bionics, year of manufacture 2014) or Locomat Pro (model LO218 by Hocoma AG, year of manufacture 2014) and DPT. The dynamic parapodium is a dynamic, individualized uprighting device – a combination of thoraco-lumbo-sacral orthosis (TLSO) and hip-knee-ankle-foot orthosis (HKAFO) – that allows the patient to stand and simulate walking by swinging the trunk. This type of gait rehabilitation (DPT) is still used in Poland in SCI patient rehabilitation centers and at home.

In addition, both groups had physiotherapy training: 1 hour of exercise according to physiotherapeutic methods and 1 hour of proprioceptive neuromuscular facilitation (PNF) conducted by a physiotherapist certified in PNF. Patients with certain medical conditions (described below) underwent complementary therapy, including classical massage, hydromassage, laser therapy, or dry CO₂ baths. Hydromassage and classic muscle massage (15–30 min) were performed to reduce muscle tension; electrostimulation (2–20 Hz, 20 min) was used to strengthen the muscles of the lower limbs; laser therapy (808 nm, 4.0 J/cm², 5–10 min) was used to reduce inflammation of tendons, fascia, and tendon sheaths; and a dry carbonic acid bath (10 min) was used to improve venous and lymphatic circulation in patients with edema.

The RAGT and DPT training sessions, along with complementary therapy, were conducted with the assistance of experienced physiotherapy staff. Each physiotherapy or complementary therapy session was documented in the treatment chart by the physiotherapist.

Patients participating in the study underwent regular medical examinations twice a week, as well as daily nursing and physiotherapy observations. Any adverse events (eg, skin abrasions) were recorded in the medical records. If such events occurred, appropriate medical intervention was initiated.

URODYNAMIC EXAMINATION:

In all patients, at the beginning (before commencing RAGT or DPT) and 1 day after the conclusion of the 7-week therapy, urodynamic assessment was performed by an experienced urologist. In all patients, the urine culture was negative prior to the procedure.

The urodynamic examinations were conducted using Mediwatch Duet Logic G2 equipment. A 5F-gauge catheter was inserted into the bladder, an electrode for external sphincter activity measurements was taped to the inner surface of the thigh, and a catheter was inserted into the rectum for intra-abdominal pressure measurements. The bladder was filled with 50 ml/min of saline solution. During the filling phase, any abnormalities, such as uncontrolled additional contractions, patient-reported additional contractions, and episodes of incontinence, were recorded. After reaching critical bladder filling, the subject urinated into a special device that measured urethral flow. Continuous recording of external sphincter activity was carried out during both phases. We compared the amount of fluid filled into the bladder and expelled by the urethra using postvoiding residual volume.

During the filling phase, we analyzed bladder function parameters, including cystometric bladder capacity (CBC), first, normal and strong desire to void (FD, ND, and SD), and compliance. Bladder compliance was defined as the ratio of a change in bladder volume to the associated change in intravesical pressure [19]. Compliance was assessed as the distensibility of the bladder, which defines the relationship between the change in bladder volume and the change in detrusor pressure. Compliance reflects the amount of fluid in the bladder required to increase bladder pressure by 1 cmH2O (mL per cmH2O). During the voiding phase, the following parameters were assessed: urinary flow (maximum and average flow rate), flow time, bladder voided volume (VV), post-void residual volume (PRV), and maximum detrusor pressure (Max Pdet).

Detrusor overactivity (DO) was defined as a urodynamic observation characterized by involuntary detrusor contractions during the filling phase, which can be provoked or spontaneous. During the voiding phase, external sphincter continuous or intermittent contractions were defined as detrusor external sphincter dyssynergia (DSD).

STATISTICAL METHODS:

Continuous variables are presented as medians and 1^st and 3^rd quartiles. Nominal variables are presented as the number of cases and the percentage. The comparison of urodynamic parameters before and after rehabilitation was performed using the non-parametric Wilcoxon signed rank test for paired data. The comparison of parameters between 2 groups was performed using the non-parametric Mann-Whitney test. For analyses comparing more than 3 groups, the non-parametric Kruskal-Wallis analysis was used. To assess the effect of the group on the change in the urodynamic parameters while adjusting for its baseline value, a rank-based ANCOVA was conducted due to the violation of assumptions for a parametric ANCOVA. Ranks were computed for both the change in values (rank_diff) and baseline values (rank_baseline), and a linear model was fitted.

The analysis was performed in the R program (R Core Team, 2021, R Foundation for Statistical Computing, Vienna, Austria). The plots were generated in the R program with the help of the ‘ggstatsplot’ ‘ggplot2’ and ‘cowplot’ package [20,21]. Tables were generated using the ‘qwraps2’ package [22].

Results

CHARACTERISTICS OF THE INVESTIGATED GROUPS:

Initially, 105 patients with SCI met the inclusion criteria and urodynamics was assessed in 51 patients. Due to secondary problems, the number of patients included in the study was markedly reduced (Figure 1). In 35 (26 males and 9 females) participants out of 51, urodynamics was assessed twice (eg, due to loss of data resulting from lack of consent for the second examination, urinary tract infections). Out of 35 patients, 10 underwent S0 rehabilitation and 25 underwent S1 (RAGT) rehabilitation. A detailed study flowchart is shown in Figure 1. Urodynamic examinations were performed for 51 patients with various types of bladder disorders (Table 1).

All patients recruited to the study had neurological impairment assessed – 12 (34.3%) had American Spinal Injury Association (ASIA) Impairment Scale (AIS) scores of AIS-A, and 23 (65.7%) had AIS-B, C, or D. The characteristics of the S1 and S0 groups are presented in Table 2.

COMPARISON OF EFFECTIVENESS OF REHABILITATION S1 VERSUS S0 GROUP:

Due to the heterogeneous clinical stages of SCI (different severity of SCI), the values of bladder parameters already differed between the groups (the S0 and S1 groups) at the beginning of therapy (Figure 2).

Although the differences were not statistically significant, patients in the S0 group had higher CBC (median S0=443 ml vs S1=294 ml), FD (median S0=269.50 ml vs S1=184 ml), ND (median S0=345.50 ml vs S1 206 ml), SD (median S0=375 ml vs S1 257 ml,) and compliance (median S0=36.90 ml vs S1 6.85 ml). At the beginning of the study, the S0 group had lower maximum flow rate (median S0=6.75 [ml/s] vs S1=10.30 [ml/s]) and average flow rate (median S0=2.40 [ml/s] vs S1=2.80 [ml/s]), and had shorter voiding time (median S0=30.50 s vs S1=59.00 s), lower Max Pdet (median S0=73 cmH₂O vs S1=151 cmH₂O), and lower VV (median S0=88 ml vs S1=102.5 ml) compared to the S1 group.

The effects of rehabilitation on bladder function did not differ significantly between the groups (Figure 2). However, for ND and SD, the between-group differences in change showed a trend towards significance, with both comparisons yielding P=0.08 in the rank ANCOVA adjusted for baseline values (Table 3).

THE IMPACT OF S1 REHABILITATION ON CHANGES IN URODYNAMIC PARAMETERS:

Figure 3 shows changes in urodynamic parameters before and after rehabilitation in the RAGT group.

No statistically significant differences were observed in the values of urodynamic parameters after rehabilitation in the S1 group. While no statistically significant change in micturition time was observed after rehabilitation in the S1 group (V=13, P=0.18, n=6 pairs), the data suggest a downward trend, which may indicate a clinically relevant improvement (Figure 3).

The associations of sex, age, level of injury, and time of rehabilitation program with of urodynamic changes in the S1 group were also analyzed.

EFFECTIVENESS OF REHABILITATION BETWEEN S1 PATIENTS WITH COMPLETE AND INCOMPLETE SPINAL CORD INJURY: Most patients with complete SCI had an increase in CBC, FD, ND, and SD after rehabilitation (Table 4). In patients with incomplete injury, these parameters were more variable. Compliance in patients with complete SCI decreased only in some cases (up to 3 ml/cm). In patients with incomplete SCI, flow rates (average and maximum) increased while micturition time decreased (Figure 4).

EFFECTIVENESS OF REHABILITATION IN S1 PATIENTS WITH DIFFERENT LEVELS OF INJURY: The lowest gains in parameters or reductions in CBC, FD, ND, and SD were observed in patients with lumbar SCI level (Table 5). Among these patients, an increase in average flow rate and a slight decrease in voiding time were observed after rehabilitation, but differences between groups with different levels of SCI were not statistically significant (Figure 5).

EFFECTIVENESS OF REHABILITATION BETWEEN S1 PATIENTS DIFFERING IN OTHER PARAMETERS (AGE, SEX, TIME FROM SCI TO START OF REHABILITATION):

There were no correlations between urodynamic parameters and age, sex, and time of injury.

Discussion

URODYNAMIC CHANGES AND TYPES OF REHABILITATION (RAGT VS DPT) IN PATIENTS WITH SCI:

Despite the lack of statistical significance of most urodynamic parameters in the study, there was a trend of greater improvement in the patients receiving RAGT versus DPT therapy. This may be related to the muscle activity that takes place during RAGT, while DPT training is basically based on trunk balance, with less involvement of the pelvic floor muscles.

Many SCI patients performing DPT experience fatigue [27,28] due to altered autonomic nervous system function [29] and the intense muscle strength required. DPT therapy involves excessive movement of the upper torso and limbs. As this device does not assist movement, it requires increased patient involvement. Excessive movement of the trunk and limbs during walking may result in greater energy expenditure. A half-hour set of motor exercises for therapy can be exhausting for patients undergoing DPT and less exhausting for patients undergoing RAGT.

No studies were identified in the literature that examined urodynamic assessment after DPT. In the study by Giannantoni et al, however, urodynamic parameters deteriorated when a reciprocating gait orthosis was used in patients with SCI [30]. This device functions differently from DPT or RAGT, as it is a bilateral hip-knee-ankle-foot orthosis that controls hip extension while assisting reciprocal hip flexion [30], and may also require greater energy expenditure from patients.

URODYNAMIC CHANGES IN SCI PATIENTS UNDERGOING RAGT:

Urodynamic parameters did not significantly change after rehabilitation in the S1 group, but the micturition time decreased after rehabilitation. The lack of significant differences in urodynamic parameters in patients with RAGT might be explained by the short duration of rehabilitation. The use of different RAGT systems (exoskeleton EKSO_GT vs Lokomat Pro) could also have influenced the results. Training with EKSO increases activation of the trunk muscles [31], and probably also the pelvic muscles. Pelvic floor muscle activation seems to be especially important in patients with SCI, 80% of whom have neurogenic bladder dysfunction. Pelvic floor muscle contractions can reflexively relax the detrusor muscle, effectively reducing its overactivity and increasing bladder capacity [32], but no such effect was observed during the use of Locomat Pro as RAGT [31]. Wiliams et al assessed a small, homogeneous group of 5 of the 12 included SCI patients, finding that pelvic floor muscle activity using electromyography was greater in the exoskeleton group, but the lower urinary tract function using urodynamics did not clearly change [17,18]. Despite the methodological differences, these results are consistent with our findings.

In the present study, we observed a reduction in micturition duration (voiding time). If voiding is completed without interruption, voiding time is equal to flow time [33]. In the Williams et al study, voiding time results were not reported. Physiologically, urinary voiding time in healthy male and female patients varies according to bladder volume, and ranges from 10 to 20 seconds for a volume of 100 ml to 25 to 35 seconds for a volume of 400 ml [34]. After SCI, this can vary [35]. This may be due to the differences in bladder size and sex, as well as differences in urodynamics due to the heterogeneous clinical picture of SCI (complete vs incomplete SCI).

Greater benefits from RAGT have been reported in patients with incomplete SCI than in those with complete SCI [8]. Studies have shown that RAGT in patients with incomplete injury (AIS B, C, or D) can lead to improvements in walking speed, distance, and functional parameters in individuals with SCI [8]. RAGT supports neuroplasticity through repetitive training, which is crucial for recovery in patients with some preserved function (incomplete SCI). However, we did not observe this trend in urodynamic changes after RAGT when comparing the complete and incomplete SCI groups. Patients with complete SCI are a more homogeneous group, which contributes to the relatively homogeneous changes of urodynamic in this group of patients after RAGT, while patients with incomplete SCI have more variable results.

The smallest lesions were observed in patients with damage from the lower levels of the spinal cord (lumbar segment). Lumbar SCI can be more detrimental to bladder function than thoracic or cervical SCI due to the location of the micturition center (S2–S4, parasympathetic system) and its impact on bladder reflexes, disrupting the reflex arc that controls bladder emptying [36,37]. Lumbar injuries often result in a flaccid bladder, leading to overdistension and damage [36,37].

However, despite the observed trends, the differences were not significant, indicating that bladder behavior cannot be predicted by the level and completeness of the spinal cord injury. Hence, performing urodynamic testing is recommended in SCI patients [38].

LIMITATIONS:

The main limitations are the small group of respondents and the lack of a control group comparable in number to the study group. One of the most important reasons is the fact that the study was conducted during the time of COVID-19 pandemic, which limited access to medical care units. As a result, the patients’ stays in the research center were interrupted, and the control urodynamic examination was not performed. In addition, there were other factors that hindered the collection of the study, such as patients opting out of the second urodynamics, due to patient soreness, and fear of a second test and urinary tract infections. This also made it difficult to collect more patients with a two-point rating. Moreover, not all urodynamic parameters were determined in the urodynamic tests. The extent of missing urodynamic data was a major limitation of the study, precluding the use of imputation and restricting the analysis to observed values only. In addition, because some of the patients assigned to the group without RAGT withdrew from the study, there were fewer people allocated to the control group. All patients wanted to be classified in the group with RAGT. These patients were not included in the study. Data from those patients were not presented in the analysis.

A limitation of the study lies in the randomization process. ‘Tossing a coin’ is a simple method of random selection used to divide participants into 2 groups; however, it may be underpowered for small samples.

Although the patients came from different parts of the country, this study was conducted in one center. Undoubtedly, a deeper analysis of the described problem would require multicenter or international studies.

In addition, the fact that the values of urodynamic parameters already differed between the S0 and S1 groups before rehabilitation may distort the results of the analysis of rehabilitation efficacy. This was due to the varied clinical condition of various SCI levels and the different levels of bladder damage that SCI can produce.

The authors of the study did not analyze all the factors and correlations of the possible influence of various parameters on the results of the study. The inclusion of these parameters could undoubtedly put the results of this study in a different light.

Conclusions

The study shows a trend toward improved of bladder function in favor of RAGT. However, the observed changes in urodynamic parameters vary among patients with different levels and degrees of completeness of SCI and show no statistical correlation. Therefore, there is a need for further research and development of more individualized and integrated technologies to support functional mobility in SCI patients and to evaluate their impact on bladder function. Further research on RAGT will help determine the optimal parameters for RAGT (eg, duration, intensity), fully understand the long-term effects of RAGT on different SCI populations and ensure its integration into clinical practice.