Benefits of the Genium microprocessor controlled prosthetic knee on ambulation, mobility, activities of daily living and quality of life: a systematic literature review

Abstract

Purpose

The benefits of advanced hydraulic microprocessor controlled knee (MPK) joints have been well established and repeatedly confirmed. The Genium knee was introduced in 2011 containing an enhanced control concept including additional sensors and improved algorithms enabling a range of new functions for transfemoral amputees (TFAs). A systematic review was conducted to evaluate the effect of the Genium knee on ambulation, mobility, activities of daily living (ADLs) and quality of life compared to standard MPKs.

Materials and Methods

The review was conducted according to PRISMA Guidelines and Recommendations of the State-of-Science Evidence Report Guidelines of the American Academy of Orthotists & Prosthetists.

Results

Twelve articles were included in the review and reported primarily on active subjects (MFCL-3/4) transitioning from C-Leg to Genium knee systems. The overall validity of the evidence was medium to high with the exception of one article having low validity. Synthesis of biomechanical analyses concludes that gait during level walking, stairs and ramps is more physiological and symmetric following accommodation and use of the Genium in community ambulating TFAs. Further, sound side loading and compensatory motions are reduced. Transitioning from C-Leg to the Genium knee additionally resulted in significant improvements in mobility, quality of life and the performance in activities of daily living (ADLs).

Conclusion

A high level of evidence was identified when assessing the ability of Genium to improve gait quality and safety and performance in ADLs. While individual studies report significant improvements in terms of quality of life and mobility, additional studies are needed to increase the evidence level.

Implications for rehabilitation

• Microprocessor controlled prosthetic knees (MPKs) are well-established devices to serve patients with transfemoral amputation. Studies conducted mostly with the C-Leg MPK show that such knees significantly increase patient safety, ambulation, mobility, performance in activities of daily living and quality of life.

• Genium MPK includes advanced features which enable a range of new functions and functional benefits to patients. Transitioning from conventional MPKs (i.e., C-Leg) to Genium MPK resulted in more physiological gait, more equally distributed loading between the prosthetic and sound limbs, as well as reduced compensatory movements on the sounds side. These outcomes could potentially reduce the long-term risks of secondary physical complications in prosthetic users (i.e., osteoarthritis, osteoporosis). Genium significantly improved mobility, performance in activities of daily living, and quality of life in the patients using a conventional MPK (C-Leg).

• Different functioning principles of the MPKs presently available are responsible for different performance levels the knees offer to users. The amount of clinical evidence is also knee-dependent, with the C-Leg knee being most extensively tested in clinical studies. This systematic review concludes that Genium offers further advantages to transfemoral patients as compared to conventional MPKs (C-Leg).

Keywords:

• Above-knee
• amputee
• Genium
• microprocessor-controlled knee
• prosthesis
• transfemoral
• X2

Introduction

A person living with transfemoral amputation (TFA) faces distinct challenges, such as increased ambulatory energy requirements, balance problems, and limitations in the ability to perform activities of daily living (ADLs) [1–3]. Adequate selection of prosthetic components, especially the knee, is of high importance to ensure optimal rehabilitation outcomes. In addition to restoring physiological biomechanics, the knee must provide maximum safety and stability. Microprocessor controlled prosthetic knees (MPKs) are well-established devices to serve patients with TFA. Different functioning principles of the MPKs presently available (i.e., hydraulic units, magnetorheologic fluid) are responsible for different performance levels these knees offer to users [4–8]. The C-Leg (Otto Bock Healthcare, Duderstadt, Germany) is a hydraulic MPK and the most researched prosthetic device in clinical studies to date with 55 publications reporting safety and performance outcomes compared to non-microprocessor controlled knees (NMPK) [9–27]. In contrast to most other MPKs, the C-Leg utilizes a default stance setting, which means that it always offers high stance flexion resistance to support the body weight unless its stance control is disengaged, reflecting its highest priority to provide safety [5].

Perhaps the most clinically relevant benefit of MPKs (C-Leg) is increased safety. Published literature suggests that falling is a major challenge for TFAs with recurrent falls and fear of falling leading to long-term functional deterioration, activity avoidance, decreased mobility and independence [28,29]. Research on MPKs, in particular on C-Leg, suggests that it reduces the number of falls, stumbles and fear of falling by up to 85%, 59% and 65.5%, respectively [9–14]. Significant improvements have also been reported for validated clinical instruments such as the Berg Balance Scale or the Timed up and Go test [10,15,16]. The effect of MPKs (i.e., C-Leg) on ambulation has been extensively studied by means of biomechanical analysis and assessments during level walking and various obstacle parcours have reported an increase in walking velocity [11,12,17–21]. Improved gait symmetry and gait pattern harmonization have also been reported as well as more physiological swing phase control during varying speeds [14,18,22]. Transitioning from a NMPK to the MPK resulted in faster and more symmetrical gait during ramp ascent and descent [9,12,14,23] but also improved overall stair mobility with the knee enabling increased loading on the prosthetic side during step-over-step descent [11,12,24]. Significant improvements in mobility have been shown in both limited (MFCL-2) and unlimited (MFCL-3) community ambulators [11,12]. The cumulative effect of the aforementioned benefits of MPKs contribute to improved performance in ADLs and better quality of life [11,12,25,26].

A new MPK, Genium (Otto Bock Healthcare Duderstadt, Germany), was introduced in 2011 and is also, similar to the C-Leg, a hydraulic default stance MPK. Advanced features (i.e., new sensors and algorithms) enable the knee to offer a range of new functions and functional benefits to patients [30]. A new core function of the knee is called “Optimized Physiological Gait.” This contains several novel features supporting more physiological gait such as four degrees of pre-flexion at initial contact which purportedly promotes increased stance flexion and shock absorption, reciprocal ramp and stair descent. Additionally, adaptive yielding control assists with stance flexion and extension, dynamic stability control ensures appropriate timing of swing release in different directions, while swing phase uses adaptive control to provide speed-independent knee flexion for proper toe clearance. Furthermore, there is a new function to support ascending stairs and crossing over obstacles, a walk to run function, a stance function that blocks knee flexion when the user is standing still, second modes that could be used for programing of other activities (MyModes), and saltwater-resistance which is included in the Genium X3 version. The knee also features a novel, user-friendly software (X-Soft) which guides prosthetists through the alignment, adjustment and coordination of the entire system in a step by step manner. The Computer Assisted Alignment feature provides a wireless read-out of the Genium’s sensors in real-time and provides a graphical representation of force application points and force vectors to facilitate correct alignment [30,31].

It is currently unclear if the novel features of the Genium MPK system translate into improved clinical function of the patient. Therefore, the objective of this systematic review was to analyze the available evidence on the effects of the Genium knee on ambulation, mobility, ADL performance and quality of life in patients with TFA. In particular, it was of interest to assess potential additional benefits it may provide compared to currently available conventional MPK.

Materials and methods

The systematic review was conducted according to the PRISMA Guidelines and following the recommendations of the State-of-Science Evidence Report Guidelines of the American Academy of Orthotists & Prosthetists [32,33].

The databases PubMed, Cinahl and Cochrane Database of Systematic Reviews were electronically searched for eligible studies on January 18th, 2017. Search terms used were: (Genium) OR (knee AND microprocessor AND control*) OR X2 knee. Term “X2 knee” was used in the search because it refers to the product name used during the final stages of Genium development. It is equivalent in terms of safety and performance to the current Genium knee which was subsequently introduced to the market. The results were limited to journal articles and publications since 2012. Pertinence assessment was performed by screening the titles and abstracts in a first step and reading the full-text article in a second step if pertinence was unclear for inclusion criteria by one reviewer (L. R.). In case of uncertainty, a second reviewer was consulted (M. P. M.). Publications were eligible if they were published in the English language and included subjects with transfemoral amputation or knee disarticulation. Included were studies that compared ambulation, mobility, ADLs and quality of life with the Genium or X2 prosthetic knee to any other non-microprocessor controlled knee (NMPK) or MPK. Single-case studies were excluded. Additionally, the references of identified articles were screened for additional relevant literature.

A data extraction sheet was developed. Refinements were made according to the applicability of the form. Data extracted from each included study were the following: (1) characteristics of study participants (amputation level, aetiology of the amputation, mean age, mean time since amputation, activity level); (2) study design (intervention, randomization, duration of treatment, accommodation period); and (3) outcome (type of outcome measure, results). The risk of bias was evaluated by three reviewers (M. P. M., L. R., and M. J. H.) independently by applying the State-of-Science Evidence Report Guidelines of the American Academy of Orthotists & Prosthetists [33]. The study design was classified according to the Study Design Classification Scale [33] (Table 1). The risk of bias (and therefore the quality) of the included studies was assessed in terms of 18 potential threats to internal and 8 potential threats to external validity. Items were classified as not applicable to the design of prosthetic studies in some cases. Publications were classified as having “high” validity if they met >80% of criteria, “moderate” if they met 60–80% of criteria, and “low” if they met <60% of the criteria [45]. Internal and external validity was rated separately while the lower of the two was used in defining overall validity. The results of the reviewers were compared and discussed in case of disagreement to reach a consensus. In cases where consensus was not reached, the lead evidence report author (M. P. M.) reached the final decision.

Table 1. Body of evidence.

		Internal validity																				External validity
	Study design	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	# threats	Overall IV	1	2	3	4	5	6	7	8	# threats	Overall EV	Overall V
Hahn et al. [34]	O3	na	na	na	na	na		•	•		•	na	na		•	•	•	•		4	Moderate	•	•	•		•	•	•	•	1	High	Moderate
Highsmith et al. [35]	E1	•	•	•	•	•		•	•		•	•	•	•	•	•	•	•	•	2	High	•	•	•	•	•	•	•	•	0	High	High
Highsmith et al. [36]	E1	•	•	•	•	•		•		•	•	•	•	•	•	•	•	•	•	2	High	•	•	•	•	•	•	•	•	0	High	High
Highsmith et al. [30]	E1	•	•	•	•	•		•	•		•	•	•	•	•	•	•	•	•	2	High	•	•	•	•	•	•	•	•	0	High	High
Bell et al. [37]	E5	na	na	na	na	•		•	•		•			•	•		•	•	•	5	Moderate	•	•	•	•	•	•	•	•	0	High	Moderate
Lura et al. [38]	E1	•	•	•	•	•		•		•	•	•	•	•	•		•	•	•	3	High	•	•	•	•	•	•	•	•	0	High	High
Aldridge Whitehead et al. [39]	E5	na	na	na	na	•		•	•		•			•	•		•	•	•	5	Moderate	•	•	•	•	•	•	•	•	0	High	Moderate
Highsmith et al. [40]	E1	•	•	•	•	•		•	•		•	•		•	•		•	•	•	4	Moderate	•	•	•	•	•	•	•	•	0	High	Moderate
Highsmith et al. [41]	E1	•	•	•	•	•		•	•		•	•	•	•	•		•	•	•	3	High	•	•	•	•	•	•	•	•	0	High	High
Kannenberg et al. [42]	E5	na	na	na	na	•				na	•				•		•	•		8	Low	•	•	•		•	•	•	•	0	High	Low
Bellmann et al. [43]	E5	na	na	na	na	•		•	•		•			•	•		•	•		6	Moderate	•	•	•	•	•	•	•	•	0	High	Moderate
Bellmann et al. [44]	E5	na	na	na	na	•		•	•		•			•	•		•	•		6	Moderate	•	•	•	•	•	•	•	•	0	High	Moderate

Study design classification: S1 – meta-analysis, S2 – systematic review, E1 – randomized controlled trial, E2 – controlled trial, E3 – interrupted time series trial, E4 – single subject experimental trial, E5 – before-and-after trial, O1 – cohort study, O2 – case–control study, O3 – cross-sectional study, O4 – qualitative study, O5 – case series, X1 – group consensus, X2 – individual opinion. Internal treats: comparison or control group used (IV 1), groups formed by random assignment (IV 2), groups comparable at baseline (IV 3), groups handled the same way (IV 4), control/comparison group appropriate (IV 5), intervention(s) blinded (IV 6), inclusion criteria appropriate (IV 7), exclusion criteria appropriate (IV 8), protocol addresses fatigue and learning (IV 9), protocol addresses accommodation and washout (IV 10), attrition explained and less than 20% (IV 11), attrition equal between groups (IV 12), outcome measures reliable (IV 13), statistical analysis appropriate (IV 14), effect size reported (IV 15), statistical significance reported (IV 16), statistical power adequate (IV 17) and free from conflicts of interest (IV 18). External treats: sample characteristics adequately described (EV 1), sample representative of the target population (EV 2), outcome measures adequately described (EV 3), outcome measures valid for this study (EV 4), intervention adequately described (EV 5), findings clinically significant/relevant (EV 6), conclusions placed in context of existing literature (EV 7) and conclusions supported by findings (EV 8). “na” is a non-applicable criteria for this study type; “•” indicates fulfilled criteria; blank spaces indicate threats to the validity (criteria not fulfilled).

The included publications were used to develop the evidence statements (ESs) [33]. They were grouped into four outcome topics (ambulation, mobility, activities of daily living and quality of life), summarizing the statistically significant findings. The ESs were developed if one or more publications reported results pertaining to a single outcome group. The ESs were then evaluated by examining: the number of publications contributing to the ES, the overall methodological quality of contributing studies and whether the findings were consistent or conflicting. A level of evidence for the individual ESs was defined as “high,” “moderate,” “low” or “insufficient.” “High” evidence level was assigned to the ES when multiple peer-reviewed publications of moderate to high overall methodological quality reported consistent results in support of the statement. A “moderate” evidence level was assigned when multiple peer-reviewed publications of low to moderate overall methodological quality reported consistent results. A “low” evidence level was assigned when there was a single peer-reviewed publication of moderate to high overall methodological quality, or multiple peer-reviewed studies of low methodological quality or if substantial findings among studies were inconsistent or conflicting. Finally, “inconsistent” was used to describe insufficient evidence to support the statement due to multiple studies with conflicting results.

Results

Study selection, characteristics and risk of bias

The database search identified 57 articles (Figure 1). Additionally, two articles were identified through reference searching. The titles and abstracts of 51 articles were screened, 34 of which were excluded for obvious non-pertinence. Seventeen full-text articles were assessed for pertinence. After the elimination of duplicates and non-eligible articles, 12 publications remained for detailed analysis in this review (Table 2).

Figure 1. Flowchart of literature search and analysis.

Table 2. Study characteristics.

	Number of subjects	Mean age (years ± SD)	Aetiology	Amputation level	Mean time since amputation	Activity level	Study design	Mean follow-up period	Products
Hahn et al. [34]	899	49.0 ± 12.9	Trauma (68.9%), tumour (15.4%), vascular disease (6%), other (9.7%)	TF (80.1%) and KD (18.9%)	First prosthesis since 21.2 ± 15.6 years	K 2 (12.5%), K 3 (64.1%), K 4 (22.8%)	Cross-sectional	1 week	Genium versus C-Leg (77%), various mechanical (hydrau\lic, brake, etc.) (23%)
Highsmith et al. [35]	20	46.5 ± 14.2	70% trauma, 20% malignancy, 10% vascular	TF	17.7 ± 15.6 years	K 3	RCT	3 months	Genium versus C-Leg
Highsmith et al. [36]	20	46.5 ± 14.2	70% trauma, 20% malignancy, 10% vascular	TF	17.7 ± 15.6 years	K 3	RCT	3 months	Genium versus C-Leg
Highsmith et al. [30]	25 (20 amp, 5 nonamp)	46.5 ± 14.2 (AMP); 57.2 ± 15.7 (NAMP)	75% Trauma, 20% malignancy, 5% PVD	TF	17.7 ± 15.6 years	K 3	RCT	3 months	Genium versus C-Leg
Bell et al. [37]	21	32.7 ± 5.3	Trauma	TF	More than 2 years	K 4	CBAT	112 days	X2 versus C-Leg, Rheo knee, total knee, Mauch
Lura et al. [38]	20	46.5 ± 14.2	70% trauma, 20% malignancy, 10% vascular	TF	17.7 ± 15.6 years	K 3	RCT	3 months	Genium versus C-Leg
Aldridge Whitehead et al. [39]	14	31.1	Not reported	TF	More than 6 months	Minimum K 3	CBAT	4 months	X2 versus C-Leg (12) and total knee (2)
Highsmith et al. [40]	20	46.5 ± 14.2	70% trauma, 20% malignancy, 10% vascular	TF	17.7 ± 15.6 years	Minimum K 3	RCT	3 months	Genium versus C-Leg
Highsmith et al. [41]	25 (20 amp, 5 nonamp)	46.5 ± 14.2	70% trauma, 20% malignancy, 10% vascular	TF	17.7 ± 15.6 years	Minimum K 3	RCT	3 months	Genium versus C-Leg
Kannenberg et al. [42]	10	36.7 ± 10.2	Trauma	TF	12.5 ± 9.6 years	40% K3, 60% K4	CBAT	90 d	Genium versus C-Leg
Bellmann et al. [43]	11	37 ± 10.2	91% trauma, 9% tumour	TF	12.5 years (range from 3 to 34 years)	45% K3, 55% K4	CBAT	2 d	Genium versus C-Leg
Bellmann et al. [44]	10	36 (range 22 − 54)	90% trauma, 10% tumour	TF	10.4 years (range from 3 to 24 yrs)	40% K3, 60% K4	CBAT	3 months	Genium versus C-Leg

TF: transfemoral amputation; KD: knee disarticulation; RCT: randomized controlled trial; CBAT: controlled before-and-after trial.

Six publications utilized randomized controlled cross-over design [30,35,36,38,40,41], five publications before-and-after design [37,39,42–44] and one study used a cross-sectional design [34]. In many of the publications, the sample sizes were between 10 and 25, with the exception of one cross-sectional study that reported data of 899 subjects. Follow-up duration ranged across the publications from 2 days to 3 months. All the studies except for one reported on subjects with the mobility grades MFCL-3 and MFCL-4 as defined by the Medicare Function Classification Level. All subjects were long-term prosthetic users and trauma was the most common aetiology of the amputation. All but two included studies described subjects previously fitted with C-Leg transitioning to Genium.

Bias risk was assessed by evaluating internal and external validity (Table 1). In terms of internal validity, the number of studies scoring high, moderate and low was 5, 6 and 1, respectively. None of the studies used blinding (Internal validity (IV) criteria 6) which represents a threat. Additional common threats to the internal validity observed in included studies were: lack of the protocol addressing fatigue and learning (IV 9), no effect size reported (IV 15), attrition not explained and more than 20% (IV 11), attrition not equal between groups (IV 12) and not free of conflict of interest (IV 18). External validity was ranked as high for all the included studies. Finally, when considering both internal and external validity, there were five studies of high, six of moderate and one of low overall validity.

Ambulation

Nine articles focused on ambulation, in particular on level walking, stairs and ramps (Table 3) [34–39,41,43,44].

Table 3. Summary of evidence.

Study	Overall validity	Outcomes	Results
Ambulation
Level walking
Hahn et al. [34]	Moderate	Prosthetist assessed functional benefits and advanced maneuvers	Responsiveness^+ in functional benefits with Genium in range from 95% and 97% (i.e., symmetry, relief, attention, variable speed, effort. Responsiveness in advanced manoeuvres with Genium: small steps (74.94%), walking backwards (91.28%), obstacles (75.32%)
		Patient perceived benefit	Responsiveness^+ in perceived benefit with Genium in range of 85–90% (obstacle, variable speed, small step, walking backwards, visual obstruction)
Lura et al. [38]	High	Motion analysis during level walking and additional weight	Increase in peak knee flexion angle in swing phase with Genium* (5–7°; except for fast with 2° increase) and stance phase* (1–2°, only slow and normal significant). During weighted walk trial, increase in peak knee flexion angle with Genium in swing phase* (3–6°; all speeds) and stance phase* (2–30). Knee flexion angles with Genium closer to control values
Bellmann et al. [43]	Moderate	Walking speed	No significant difference. Slow walking velocity (mean ± SD): 1.02 ± 0.14 m/s C-Leg, 1.09 ± 0.12 m/s Genium Medium walking velocity (mean ± SD): 1.29 ± 0.14 m/s C-Leg, 1.29 ± 0.13 m/s Genium Fast walking velocity (mean ± SD): 1.57 ± 0.18 m/s C-Leg, 1.53 ± 0.25 m/s Genium
		Motion analysis at 3 different speeds	Asymmetry of step length for slow and medium velocities decreased* with Genium: Step length asymmetry at slow speed (mean ± SD): 0.05 ± 0.06 m C-Leg, 0.02* ± 0.06 m Genium Step length asymmetry at medium speed (mean ± SD): 0.07 ± 0.05 m C-Leg, 0.03* ± 0.07 m Genium Step length asymmetry at fast speed (mean ± SD): 0.08 ± 0.04 m C-Leg, 0.05 ± 0.07m Genium Peak knee flexion angle in swing phase nearly constant with Genium (63.8°) and not speed dependent. With C-Leg maximum knee flexion angle increase by 14.6° for 1m/s increase in speed. Peak swing knee flexion angle during small steps increased* with Genium (mean maximum knee flexion angle was 46.1° with Genium versus 41.7° with C-Leg). Switching into swing phase mode during small steps correctly done for 95.1% steps with Genium versus 75.3% with C-Leg. 4° increase in knee flexion at initial contact with Genium than with C-Leg. There are differences between subjects in performing an increased stance phase flexion (subgroup of those using increased stance phase flexion with Genium increased). Lower vertical and horizontal ground reaction forces at weight acceptance on prosthetic side with Genium. Increased knee flexion moments (significant only for fast speed) with Genium.
Highsmith et al. [41]	High	Walking velocity on even and uneven terrain	Similar results for C-Leg and Genium (no significant difference). Non-amputees completed all the tests faster* than amputees, regardless of knee (C-Leg or Genium)
		Borg’s rate of perceived exertion (RPE)	RPE for nonamputees was higher than that of the amputee group regardless of knee
		Postural stability (PS) and limits of stability (LOS)	PS: No difference between C-Leg and Genium. LOS: C-Leg offered improved anterolateral directional control* compared to Genium. Genium offered improved control in rearward direction* compared to C-Leg. Non-amputees had better time to complete and control anteriorly, in the direction of the involved side (also anterolaterally and posterolaterally).
Ramps
Hahn et al. [34]	Moderate	Prosthetist assessed advanced maneuvers	Responsiveness^+ in ramp descent (70.86%), ramp ascent (74.53%)
		Patient perceived benefit	Responsiveness^+ in perceived benefit in range of 85–95% (standing slope, slope down, slope up)
Lura et al. [38]	High	Motion analysis on 5° and 10° ramp	Increased prosthetic side peak swing knee flexion angle during 5° ramp ascent* (3–8°, all 3 speeds), and during descent* (9°, only normal speed significant) when transitioning from C-Leg to Genium . No significant differences during ascent in the stance phase between knees. During descent, Genium knee flexion in stance increased* (slow 3°; fast 4°). Similar results observed on 10° ramp
Bellmann et al. [43]	Moderate	Standing on slope (loading, moments and postural sway)	Increased* loading on prosthetic side with Genium (353N) than C-Leg (190 N). Increased* sagittal external flexion moment of the prosthetic knee join axis (17.5 Ncm/kg Genium versus 9.1 Ncm/kg C-Leg). Reduced* moments of the hip centre of rotation on affected side (1.4 Ncm/kg Genium versus 4.5Ncm/kg C-Leg). Sway (total distance of centre of pressure) increased* with C-Leg.
		Motion analysis during 10° ascent and descent	Increased* prosthetic knee flexion during ascent and descent with Genium (ascending: Genium 56.7°, C-Leg 49.9°; descending: Genium 72.4°, C-Leg 63.5°). During descent, prosthetic sagittal external knee flexion moments increased* (Genium −0.85 Nm/kg, C-Leg −0.76 Nm/kg). Trend of decrease in contralateral vertical ground reaction forces at weight acceptance
Bell et al. [37]	Moderate	Walking speed	Self-selected speed increased* by 0.1 m/s when transition from MPKs (C-Leg, Rheo, 1.04 m/s) to Genium (1.14 m/s)
		Hill Assessment Index	No difference between MPK (C-Leg, Rheo, score 11) group and Genium (score 11). Index increased* from 8.5 with mechanical knee to 11 with Genium
		Motion analysis during slope descent	Transitioning from MPKs to Genium: increased prosthetic knee flexion to a median 6.40°* at initial contact and 73.70 degrees* in swing phase, longer prosthetic limb step*, increased prosthetic limb impact peaks*, improved impact peak symmetry*
Highsmith et al. [36]	High	Motion analysis of 5° ramp	Knee moment DoA better with Genium for slow speed up* the ramp (ES medium) and down* the ramp (ES small), comfortable speed down the ramp* (ES large), as well as fast speed up the ramp* (ES small). No difference between Genium and C-Leg during self-selected up and fast down the ramp. Variance in peak knee flexion moment was reduced with Genium during uphill and self-selected downhill, and with C-Leg during slow and fast downhill walking
Stairs
Hahn et al. [34]	Moderate	Prosthetist assessed advanced maneuvers	Responsiveness^+ in stairs down (70.52%), stairs up alternatingly (32.81%), descent (66–70%).
		Patient perceived benefit	Responsiveness^+ in stairs down (approx. 65%), stairs up alternatingly (approx. 90%)
		Percent of subjects ascending stairs alternatingly	64.29%
Aldridge Whitehead et al. [39]	Moderate	Stair Assessment Index ascent	SAI ascent improved from 5 to 11*
		Self-selected ascent strategy	7% with C-Leg/Total Knee and 71% with Genium self-selected step-over-step
		Motion analysis	Mean increase in maximum knee flexion* (increase of 62°, Genium 68.2° versus conventional/C-Leg 2°) and hip flexion* (increase of 22°, Genium 75.7° versus conventional/C-Leg 53.8°) during swing and in knee flexion at heel strike on prosthetic side with Genium when ascending stairs. Normalized knee mechanics with Genium during stair ascent when compared to non-amputees during swing phase (88.4°, p = 0.179), but not during initial contact (65.7°, p = 0.002). Positive correlation to prosthetic limb hip power during pull-up* (r = 0.641) and push-up/early swing* (r = 0.993)
Bellmann et al. [43]	Moderate	Motion analysis during ascent and descent	Descent: increased* sagittal external knee flexion moments on the prosthetic side at midstance with Genium. Similar maximum vertical ground reaction forces on contralateral side for both knees (p = 0.59) Ascent: increased* duration of stride between step-over-step (Genium) and conventional method (C-Leg). Hip and knee range of motion on contralateral side during conventional (step-to, C-Leg) method was higher* than on the prosthetic and contralateral side during step-over-step with Genium. During step-over-step method, loading of contralateral knee and hip reduced*. Maximum residual limb extension power on prosthetic side with Genium comparable to nonamputees (p = 0.63)
Bellmann et al. [44]	Moderate	Ascent strategy	80% of subjects able to use step-over-step strategy with Genium, 20% step-to pattern. 100% of subjects used step-to method with C-Leg
		Motion analysis during ascent and descent	Duration to climb a stair increased* during step-over-step method compared to step-to. Number of stairs climbed daily is 100 stair steps. Step-to stair ascent pattern (typical for C-Leg users) results in increased* hip and knee movement on the contralateral side compared to step-over-step method (80% of Genium users use the method) and healthy group. Increased* ankle movement on contralateral side during step-over-step method compared to healthy group. Reduced contralateral knee extension power and increased ankle joint power during step-over-step compared to step-to method
Highsmith et al. [35]	High	Stair Assessment Index	Median SAI ascent score improved* from 6 with C-Leg to 11 with Genium. Mean SAI ascent score for three trails improved from 5.6, 6.0 and 5.6 with C-Leg to 9.7, 9.9 and 10.1 with Genium. Median SAI descent score improved* with Genium. Mean SAI descent score for three trails improved from 11.0, 10.3 and 11.2 with C-Leg to 12, 11.7 and 11.8 with Genium
		Time to ambulate stairs	Stepping rate during ascent higher* with C-leg (1.1 steps/) than with Genium (0.8 steps/s). Step rate during descent similar between knees (1.2 steps/s C-Leg versus 1.1 steps/s Genium)
Mobility
Highsmith et al. [35]	High	Four Square Step Test	Decreased* by 1.1 s (from mean 12.2 s with C-Leg to mean 11.1 s with Genium). ES small
		Amputee Mobility Predictor	Increased* by 2 points (from median 42 with C-Leg to median 44 with Genium)
		Step activity derived functional level (K-level)	Increased* by 0.2 points (from median 3.4 with C-Leg to median 3.6 with Genium)
Activities of daily living
Hahn et al. [34]	Moderate	Prosthetist assessed advanced manoeuvres	Responsiveness* in range of 72–82% (i.e., door test, carrying objects, obstacle and uneven terrain, toileting)
		Patient perceived benefit	Responsiveness* in range of 80–90% (heavy load, busy crowd, trolley, heavy door)
Highsmith et al. [30]	High	Continuous Scale Physical Functional Performance CS-PFP10	When comparing Genium to C-Leg, the total score improved* by 7.4% ( ES = 0.28; C-Leg 55.2, Genium 59.6). Improvement* in 3 domains: upper-body flexibility (7.0%, ES = 0.45), balance (7.6%, ES = 0.28), endurance (8.4%, ES = 0.32). Positive trend in 2 domains: upper-body strength (5.4%), Lower-body strength (8.1%). When comparing non-amputees with amputees:
			1/5 domains significantly better for non-amputees when compared to Genium users (endurance).
			4/5 domains significantly better for non-amputees when compared to C-Leg users (endurance, lower-body strength, upper-body strength, balance and coordination)
Highsmith et al. [35]	High	ADLs questionnaire	Improvements* in perceived function and safety in 3 of 5 of the domains: Family and Social Roles, Leisure Time Activities; Mobility and Transportation; Other activities
			No difference in 2 domains: Personal Care and Dressing; Health-related Exercise
Kannenberg et al. [42]	Low	ADLs questionnaire	Subjects report of clinically relevant:
			gain in safety with 27 out of 45 of ADLs (60%) ;
			decrease in difficulty with 24 out of 45 ADLs (53%)
			Improvements seen in the categories of family and social life, mobility and transportation
Quality of life
Highsmith et al. [40]	Moderate	Prosthetic Evaluation Questionnaire	Improvement* in 4 out of 9 scales: Perceived Response, Social Burden, Utility, Well-Being. Positive trend in 2 out of 9 scales: Appearance and Sounds. Unchanged 3 out of 9 scales: Ambulation, Frustration, Residual Limb Health. 7/10 items related to physical performance were improved* (comfort standing, walking in close spaces, going down stairs, walking steep hill, walking slippery surface, satisfaction when walking). 3/10 not significant: feeling off balance, satisfaction, satisfaction with training

SAI: stair assessment index; SSWS, FPWS, AM, NAM; ⁺Responsiveness: percentage of subjects showing ‘clear’ or ‘very clear’ benefit; DoA: degree of asymmetry; ES: effect size.

*Significant (p < 0.05).

Level walking

Four publications of moderate to high overall validity that investigated the effect of Genium on level walking focused on biomechanical analyses but also stability and exhaustion commonly used to asses prosthetic components [34,38,41,43]. Walking speed and perceived exertion were not significantly different between Genium and C-Leg, however numerous parameters suggest more physiological and safer gait. Peak knee flexion angle in swing and stance phase was significantly increased (p < 0.05) with Genium and closer to values of able-bodied subjects, especially during slow and medium speeds and under additional weight simulating heavier footwear [38,43]. Speed dependency of maximum knee flexion angle in swing phase that is an important determinant of toe clearance was further decreased with Genium, while the peak knee flexion during swing in small steps significantly increased [43]. Due to four degrees of pre-flexion at initial contact, lower vertical and horizontal ground reaction forces were seen at weight acceptance on the prosthetic side. Increased knee flexion moments on the prosthetic side indicating increased weight-bearing were observed across different velocities but were significant only for fast speed. Asymmetry of step length at different velocities significantly decreased [43]. While postural stability control was not different between the knees, in terms of limits of stability there were some differences with C-Leg significantly improving directional stability (p ≤ 0.05) anterolaterally over the prosthesis, whereas Genium use resulted in greater (p ≤ 0.05) posterolateral directional stability over the sound side [41].

Walking on Ramps

Five publications reported on ramp walking outcomes [34,36–38,43]. Walking speed with Genium was evaluated in two studies which reported either no change or improvement with Genium [36,37]. Peak knee flexion in swing phase was significantly increased with Genium during both ramp ascent and descent at various speeds and inclinations indicating increased toe clearance [37,38,43]. During ramp descent, Genium use resulted in significantly more flexion at initial contact and higher peak knee flexion, while peak knee flexion during ascent was not significantly different from C-Leg [37,38]. Increased loading on the prosthetic side representing increased weight-bearing during descent was found by a significant increase in sagittal external knee flexion moments, which resulted in a trend of decreased contralateral vertical ground reaction forces and, thus, unloading of the sound side, at weight acceptance [43]. Additional gait harmonization was also reported in terms of reduction of knee moment asymmetry, reduction of variance in peak knee flexion moments, increased prosthetic limb step length, increased prosthetic limb impact peaks and peak symmetry [36,37]. The ability of the knee to support standing on ramps was also assessed [43]. Average loading and, thus, weight-bearing, during 3 min of standing on a 10 degree ramp was significantly increased with Genium (prosthetic side), while hip moments on the prosthetic side were significantly reduced, representing a reduction in hip compensation, as was postural sway.

Walking on stairs

Five publications assessed the performance of subjects with Genium on stairs [34,35,39,43,44]. Studies suggest that between 64% and 80% of subjects were able to ascend stars by using a step-over-step method when using the Genium knee [34,39,44]. While being more physiological, the duration to ascend stairs significantly increased during step-over-step pattern in comparison to the common step-to pattern of TFA [35,44]. The ability to utilize the step-over-step ascent method resulted in the more physiological increase in knee and hip range of motion on the prosthetic side and decrease of compensations on the sound side [39,43,44]. Increased ankle movement on the contralateral side was also reported and comparable to able-bodied subjects (p < 0.01) [44]. During the step-over-step method, loading of the contralateral knee (p < 0.05) and hip (p = 0.33) was reduced [43]. While the maximum residual limb extension power on the prosthetic side with Genium was increased and comparable to able-bodied subjects (p = 0.63), a reduction was measured for contralateral knee extension power (p = 0.13) [43,44]. Finally, ankle joint power on the contralateral side was increased with the step-over-step method [44]. Stair assessment index (SAI) was used in two studies reporting significant qualitative improvements during stair ascent represented by an average score increase from 5 or 6, respectively, with the C-Leg (5: “without rail or assistive device, step-to”; 6: “with rail or assistive device, skipping step pattern”) to 11 with Genium (“with rail, step-over-step pattern”) [35,39]. A significant improvement was also reported for the SAI descent score, but much smaller in magnitude [35]. Duration to descend stairs was unchanged when transitioning from C-Leg to Genium [35]. Finally, while sagittal external knee flexion moments on the prosthetic side at midstance were significantly increased with Genium, vertical ground reaction forces on the contralateral side were not affected [43].

The ES was defined as “Genium results in more physiological gait, unloading and reduced compensatory mechanisms of the sound side compared to conventional MPKs.” Due to nine publications of high to moderate quality and consistent findings, a “high” evidence level was assigned to the statement.

Mobility

One article of high overall validity addressed the effect of Genium on subjects’ mobility (Table 3) [35]. Four square step test completion times were significantly reduced (small effect size). Furthermore, transitioning from C-Leg to Genium resulted in the significant increase in the Amputee Mobility Predictor (AMP) by 2 points. Highsmith et al. used StepWatch (Orthocare Innovations. Edmonds, Washington) devices to record step activity derived functional levels [35]. Following two weeks of recording of step activity, the data was uploaded into the Galileo cloud whose software generated the activity report and functional level (MFCL or “K”-level). The functional level calculation was based on multiple factors, including cadence variability, potential to ambulate, ambulation requirement and the clinician’s observation of functional level. The calculated average functional level was significantly improved from 3.4 with C-Leg to 3.6 with Genium.

The following ES was developed: “Genium improves mobility of unilateral TF amputees when compared to conventional MPKs.” Due to the existence of only one high quality publication, the evidence level “low” was assigned to the ES.

Activities of daily living

Effect of Genium use on ADL performance was reported in four publications having an overall validity ranging from low to high (Table 3) [30,34,35,42]. Scores on a validated instrument, short form Continuous Scale – Physical Functional Performance (CS-PFP10), was significantly improved when transitioning from C-Leg to Genium. Significant improvements were shown in three domains (Upper-body flexibility, Balance and Endurance) and positive trends in two domains (Upper-body strength and Lower-body strength). Able-bodied controls demonstrated significantly better CS-PFP10 scores in only one domain (Endurance) as compared to TFA using Genium but in four domains (Endurance, Lower-body strength, Upper-body strength, Balance and coordination) as compared to the same TFA using C-Leg. In two other studies, a non-validated questionnaire was used to evaluate the patient-perceived safety and difficulty of 45 ADLs with Genium as compared to C-Leg [35,42]. Both publications report significant improvements in perceived safety and difficulty, especially in activities dealing with Family and Social Role and Mobility and Transportation. Kannenberg et al. reported clinically relevant gain in safety for 27 out of 45 of ADLs (60%) and greater ease of performing 24 out of 45 ADLs (53%) [42]. Finally, prosthetists appraised that over 70% of tested Genium users showed ‘clear’ or ‘very clear’ improvement in the ability to perform advanced manoeuvres (i.e., door test, carrying objects, obstacle and uneven terrain, toileting) [34]. Similar feedback was obtained from the patients themselves, who reported benefits in performing activities such as carrying heavy loads, navigating in busy crowds, pushing a trolley or heavy door in 80–90% of cases.

The following ES was developed: “Genium use improves performance in and safety of conducting ADLs compared to conventional MPKs.” Due to three high or moderate quality publications with consistent findings, it was decided that a single low-validity publication should not be allowed to dilute the overall evidence level. Therefore, a high evidence level was assigned to the statement.

Quality of life

Quality of life was addressed in one article having moderate validity (Table 3) [40]. Significant improvements in Prosthetic Evaluation Questionnaire (PEQ) domain scores were reported after 3 months when transitioning from C-Leg to Genium. Four scales were significantly improved: Perceived Response, Social Burden, Utility, and Well-Being (all p < 0.05), while positive trends were observed in two scales (Appearance and Sounds). No difference was found in the scales of Ambulation, Frustration and Residual Limb Health. Seven out of ten PEQ items related to physical performance were significantly improved with the Genium (comfort while standing, walking in tight spaces, walking down stairs, walking steep hills, walking slippery surfaces, satisfaction with walking; p < 0.025). An ES was developed stating “Genium further improves prosthesis-related quality of life of unilateral TF amputees as compared to conventional MPKs.” Due to the availability of only one moderate-quality publication, the evidence level “low” was assigned to the ES.

Discussion

Summary of the evidence

Twelve publications reporting on TFAs (MFCL-3 and 4) predominantly transitioning from C-Leg to Genium were included in the systematic review. The overall quality of the evidence was moderate to high with the exception of one article having low methodological quality. The studies were selected and their results evaluated to assess the effects of Genium on ambulation, mobility, ADL performance and quality of life in above-knee amputees as compared to conventional MPK and NMPK.

The effect of Genium on ambulation was assessed for level walking, ramps and stairs. While there was no significant influence on the speed of conducting those activities, biomechanical findings suggest that gait kinetics and kinematics as well as several spatiotemporal parameters were more physiological with Genium than with conventional MPKs. The evidence supports more physiological knee flexion in both stance and swing phase, representing increased weight-bearing and improved toe clearance, reduction of speed-dependency of maximum knee flexion angle in swing phase, representing more consistent speed-independent toe clearance, improved symmetry of step length, as well as increased toe clearance especially during small steps and simulating heavier footwear during level walking. The new stair function enabling up to 80% of subjects to perform step-over-step stair ascent resulted in more physiological movement patterns on the prosthetic side but also significant reductions in compensatory movements on the sound side. While weight acceptance on the prosthetic side was more comfortable as shown by reduced ground reaction forces, the flexion moments on the hip and knee suggest greater prosthetic side loading, i.e., weight-bearing. This finding was statistically significant for fast walking. Gait analyses on ramps showed increased maximum knee flexion angle and, thus, foot clearance similar to those measured on the sound side. The new standing function resulted in significantly more loading and, thus, weight-bearing, on the prosthetic side during standing on ramps.

The reported improvements in kinetic gait parameters as well as loading symmetry, representing a more even distribution of the body weight with Genium may be clinically relevant. Published evidence on the amputee population suggests that asymmetry, especially in loading is related to long-term consequences to the locomotor system resulting in secondary physical complications among lower limb prosthetic users, such as knee osteoarthritis [46,47]. A recent systematic review examining the risk of degenerative changes in lower limb amputees concluded that 56% of amputees suffer from back pain, 35% of patients from osteoarthritis of the sound knee, and 15% and 20% from hip osteoarthritis on the prosthetic and sound side, respectively [47]. Reduced bone density was observed in 87% of patients in their prosthetic side hip, and all amputees exhibited residual limb muscle atrophy. The prevalence of back pain and knee osteoarthritis was shown to increase in more proximal amputations. Similar conclusions were reached in a review by Gailey et al. [46]. Asymmetry, especially in terms of loading, has been identified as a major cause for those observations. Long-term exposure to high repetitive loading forces leads to the degeneration of weight-bearing joints that can contribute to joint pain [48,49]. Burke et al. attributed gait asymmetry and increased load to the intact limb to the higher incidence of osteoarthritis in long-term prosthesis users [50]. While optimized prosthetic fitting may reduce these risks, prosthetic components such as the Genium knee, may do so as well by facilitating a more physiologic gait as demonstrated by the aggregate results of the included studies [36–39,43,44]. More symmetrical gait and more physiological, more evenly distributed loading of the sound and prosthetic sides could reduce the long-term risks of secondary physical complications in prosthetic wearers.

Validated clinical instruments, such as the AMP and MFCL-level derived from StepWatch, recordings were used to assess mobility with Genium. Significant improvements were observed. However, conclusions regarding clinical relevance cannot be drawn due to lack of established reference values for the lower limb amputee population. A validated instrument assessing ADL performance further suggested that Genium reduced the functional gap between able-bodied subjects and prosthetic users. It is interesting that significant improvements were observed in upper-body flexibility that researchers attributed to greater lower body stability and reduced necessary concentration on lower extremity function as the foundation for performing upper extremity activities. Significant improvements in both perceived endurance and balance measured by a non-validated ADL questionnaire were supported by the same performance-based validated instrument (CS-PFP10). Not a single item of 45 activities was rated easier or safer to perform by subjects when done with a conventional MPK. Many activities were improved with Genium both in terms of safety and ease of execution. Prosthesis-related quality of life as measured by the PEQ was significantly improved when transitioning from C-Leg to Genium, particularly in the subscales of Perceived Response, Social Burden, Utility and Well-Being. Finally, approximately 80% of previous C-Leg users preferred Genium after training with and using it for approximately 3 months.

The reviewed results were used to develop four ES. The statements related to the ability of Genium to improve gait quality, and that regarding safety of and performance in ADL execution were evaluated as having “high” evidence levels. The other two statements assessing the ability of Genium to affect quality of life and mobility were rated as “low evidence level” due to the availability of a single high or moderate quality study only. Therefore, additional studies are needed to further corroborate the evidence of these two statements.

It should also be noted that in most studies subjects were specifically trained in the use of the Genium features and assessed for accommodation prior to assessing outcomes. One study reported for example that, when using the Genium, subjects required 0.7 ± 1.0 additional visits (0–4) for post-fitting prosthetic adjustments. Subjects required additional 3.0 ± 1.8 visits (1–8) with the study physical therapist for functional training to master new features with the Genium knee, including stance locking (i.e., the so-called intuitive stance), reciprocal stair ascent, obstacle crossing, and stance flexion feature. Subjects required 67.9 ± 27.1 d (18–119 d) to successfully complete the accommodation test with the Genium knee [30]. Results and outcomes reported in that studies are the result of the reported accommodation and training. Outcomes observed in the present review may not be generalizable in cases where training was not provided.

Limitations

Several systematic reviews conducted so far on various prosthetic components concluded that evidence in prosthetic research is of mostly low to moderate quality [51,52]. While very much true for older studies, it should be acknowledged that several attempts were made in recent years to improve the quality of studies in prosthetics. This trend is already apparent in the evidence reviewed in this systematic review with the oldest article published in 2012, resulting in most articles having moderate to high quality. Randomization (of the order of interventions) was implemented in several reviewed studies to reduce bias. On the other hand, the absence of blinding was still observed in all included studies. While subject blinding may be difficult in prosthetic research, blinding of assessors should be possible in future studies. The acclimation period when transitioning from MPKs (predominantly hydraulic knees) to Genium in most publications was three months or longer. This was considered sufficient as studies assessing transitioning from NMPKs to MPKs report similar duration of acclimation [9,53]. The acclimation period in a few studies was only two days. While this period is short, the conclusions reached in the studies were similar to those reached in studies having longer acclimation periods [37,38].

The sample size in most publications was between 10 and 25 subjects with the exception of a larger cross-sectional study, yet samples were sufficient to obtain several significant results. Internal validity of the reviewed studies could be further improved by including information addressing fatigue and learning, effect size calculation and reporting attrition rate as well as conflict of interest.

The order in which the effect of prosthetic knees was assessed varied across studies. In all the studies, subjects entering the study were using the C-Leg (or NMPK). Five publications that utilized a controlled before-and-after design assessed the patients with the C-Leg (or NMPK) prior to fitting and subsequent assessment with Genium [37,39,42–44]. In six publications where a randomized order controlled design was used, the C-Leg accommodated patients were randomized to either the C-Leg or Genium and allowed to accommodate prior to their first assessment [30,35,36,38,40,41]. After the cross-over, they were fitted with the other device prior to re-assessment. Therefore, the order in which prostheses were assessed varied and, therefore, an order effect on the results was reduced.

Conclusions

Most of the reviewed evidence reports on active, community ambulating transfemoral amputees transitioning from C-Leg to Genium. The quality of the evidence is predominantly moderate and high showing improvements in outcomes when compared to conventional MPKs and NMPKs. Accommodation with and use of Genium resulted in more physiological gait during level walking, ramp and stair ambulation, as well as reduced loading and compensatory mechanisms of the sounds side. More symmetrical gait and more physiological, evenly distributed loading of sound and prosthetic sides could perhaps contribute to reducing the long-term risks of secondary physical complications in prosthetic users (i.e., osteoarthritis, osteoporosis). Mobility and functional levels were both significantly improved, while positive effects on performance in and safety of activities of daily living were demonstrated by both performance-based and patient-reported outcome measures. Significant improvements in quality of life, as measured by the PEQ, were also reported. At this time, it is unclear if results are generalizable beyond unilaterally involved, community ambulators with TFA.

Acknowledgements

M. P. Mileusnic, L. Rettinger and M. J. Highsmith were responsible for reviewing the bias risk and for drafting the manuscript. A. Hahn provided input during drafting of the manuscript and provided critical review.

Disclosure statement

M. J. Highsmith has no financial disclosures related to preparation and drafting of this manuscript. The other authors (M. P. Mileusnic, L. Rettinger and A. Hahn) are employees of Otto Bock Healthcare Products (Vienna, Austria), the manufacturer of the MPK subject to this review. The company did not influence the design and execution of the review nor the interpretation of results. However, Ottobock allowed the authors to work on this systematic review during their regular office hours and to use the company’s office equipment (computers, software, Internet access, database access).