Anatomy of the proximal tibiofibular joint and interosseous membrane, and their contributions to joint kinematics in below-knee amputations

Abstract

A result of below-knee amputations (BKAs) is abnormal motion that occurs about the proximal tibiofibular joint (PTFJ). While it is known that joint morphology may play a role in joint kinematics, this is not well understood with respect to the PTFJ. Therefore, the purposes of this study were: (i) to characterize the anatomy of the PTFJ and statistically analyze the relationships within the joint; and (ii) to determine the relationships between the PTFJ characteristics and the degree of movement of the fibula in BKAs. The PTFJ was characterized in 40 embalmed specimens disarticulated at the knee, and amputated through the mid-tibia and fibula. Four metrics were measured: inclination angle (angle at which the fibula articulates with the tibia); tibial and fibular articular surface areas; articular surface concavity and shape. The specimens were mechanically tested by applying a load through the biceps femoris tendon, and the degree of motion about the tibiofibular joint was measured. Regression analyses were performed to determine the relationships between the different PTFJ characteristics and the magnitude of fibular abduction. Finally, Pearson correlation analyses were performed on inclination angle and surface area vs. fibular kinematics. The inclination angle measured on the fibula was significantly greater than that measured on the tibia. This difference may be attributed to differences in concavity of the tibial and fibular surfaces. Surface area measured on the tibia and fibula was not statistically different. The inclination angle was not statistically correlated to surface area. However, when correlating fibular kinematics in BKAs, inclination angle was positively correlated to the degree of fibular abduction, whereas surface area was negatively correlated. The characteristics of the PTFJ dictate the amount of fibular movement, specifically, fibular abduction in BKAs. Predicting BKA complications based on PTFJ characteristics can lead to recommendations in treatment.

Introduction

The proximal tibiofibular joint (PTFJ) has been described as a small planar type accommodatory joint (Eichenblat & Nathan, 1983) whose anatomy varies considerably across individuals (Ogden, 1974a,b). Ogden (1974a,b) was the first to describe the PTFJ using four variables: inclination angle; articular surface area; concavity and articular surface shape, which has since become the accepted nomenclature. While attempts to further characterize the PTFJ have been undertaken since the work of Ogden (1974a,b), these studies (Eichenblat & Nathan, 1983; Bozkurt et al. 2003; Espregueira-Mendes & da Silva, 2006) have produced conflicting findings and are limited in that they presented only descriptive statistics, and quantification of the relationship between joint characteristic variables was not undertaken.

The PTFJ characteristics, specifically inclination, have been implicated in both physiological and pathological movements (Ogden, 1974a,b; Sugita et al. 1995; Soavi et al. 2000; Scott et al. 2007). In physiological conditions, fibular rotation occurs to accommodate dorsi-flexion and plantar-flexion at the ankle. For example, the degree of rotation is influenced by the inclination angle such that, when the joint is horizontally inclined (opposed to oblique inclinations) the joint will experience greater external rotation during dorsi-flexion (Ogden, 1974a,b; Sugita et al. 1995). Pathologically, Ogden (1974a,b) reported a higher incidence (70%) of subluxation or dislocation in patients with oblique inclination angles when compared with horizontal inclination. Oblique inclination angles have also been shown to have smaller articular surface areas, decreasing mobility and reducing the joint's ability to accommodate large torsional stress. In contrast, the highly mobile horizontally inclined joint enables greater flexibility and therefore is more forgiving when subjected to substantial torsion forces (Ogden, 1974a,b).

A secondary component of the PTFJ is the interosseous membrane (IOM), the thin ligamentous structure that spans the majority of the space between the tibia and fibula. The IOM consists of both major (30–40 μm diameter; inferior-laterally from the tibia to fibula) and minor (0.5–1 μm diameter; no clear orientation) fibers with a 1.2 cm² proximally located foramen (Ebraheim et al. 1998). The structural arrangement of the IOM fibers provides rigid support between the two bones, and it has been reported that the IOM is capable of withstanding between 164 N and 3604 N of tensile force, depending on the direction of the applied load (Minns & Hunter, 1976). Similar to the characteristics of the PTFJ, there is very little agreement regarding the anatomical features of the IOM, in both physiological and pathological conditions.

A thorough understanding of the morphology of the major PTFJ structures is important when evaluating physiological and pathological fibular movements. Of particular interest is a clinical condition where patients with below-knee amputations (BKAs) suffer from fibular abduction, causing pain, delays in rehabilitation and ultimately complications in prosthetic fitting (Pinzur et al. 2007; Asa et al. 2014). However, the contributions of the joint and IOM anatomy to the magnitude of fibular motion have not been investigated. Therefore, the purpose of the current research was: (i) to describe the anatomy of the PTFJ surfaces, and statistically assess the relationships between the tibia and fibula components; and (ii) to assess the multivariate relationship between the anatomical characteristics of the PTFJ and IOM to fibular motion in BKAs. It was hypothesized that there would be a strong relationship between the shapes and surface areas between the tibia and fibula. It was also hypothesized that the characteristics of the PTFJ and the IOM would be significant predictors of fibular motion in BKAs.

Materials and methods

Specimen preparation

Forty embalmed specimens (22 males; 20 right) with a mean (SD) age of 77.1 (14.1) years, acquired from the Anatomy and Cell Biology Bequeathal Program at Western University, were used for this study, with permission from the Committee for Use of Cadaveric Material. The data generated here were collected in collaboration with a second study using the same specimens, described by Asa et al. (2014), where details of the dissection process can be found. Briefly, however, the specimens were disarticulated at the ankle and knee joints, and all soft tissues, aside from the IOM, PTFJ capsule and biceps femoris (BF) tendon, were removed.

Anatomical measurements

Following experimental testing (see below), the tibia and fibula were disarticulated at the PTFJ exposing the articular surfaces on the respective bones. Using a digital protractor (2-in-1 Digital Angle Rule, Shanghai, China; resolution of 0.1 °), the inclination angle of the fibula was measured by resting one arm of the protractor on the fibular articular surface while the other arm ran the length of the fibular shaft along the IOM (Eichenblat & Nathan, 1983). To calculate the angle of inclination on the fibula, 90 ° was subtracted from the angle measured. Tibial inclination was measured by drawing a line extending the IOM to the tibial plateau. The digital protractor was placed in-line with the IOM extension line and the articular surface. To calculate the angle of inclination on the tibia, the measured angle was subtracted from 90 ° (Espregueira-Mendes & da Silva, 2006).

To enable a comparison with previously published data (Espregueira-Mendes & da Silva, 2006), the joint surfaces of the tibia and fibula were assumed to be elliptical in order to calculate the joint surface areas consistently (Equation 1):

where, max and min are the maximum and minimum diameters, respectively, and were measured with digital calipers (Procise™, The Innovak Group, Montreal, Quebec< Canada; resolution: 0.01 mm; Fig.1a,b). To account for variability, PTFJ surface area measurements were averaged over three trials. The joint was further visually classified according to the following three criteria: (i) type – planar, trochoid, double-trochoid; (ii) concavity – convex, concave; and (iii) shape – circular, elliptical, square, triangle (Espregueira-Mendes & da Silva, 2006). Finally, with respect to the surface area of the IOM, it was assumed to be rectangular and for the intact IOM specimens was calculated as such. However, for statistical modeling purposes (see Data analysis and statistics section below), the surface area of the sectioned IOM specimens was assumed to be zero as the IOM provided no support as a result of sectioning (Fig.1c,d).

Figure 1.

Examples of the surface area measurements for the articular surfaces of the fibula (superior view) (a) and tibia (inferior view) (b) articular surfaces, and the IOM for a long (c) and short (d) amputation. The fibular articular surface (a) was categorized as triangular and concave, while the tibial surface (b) was categorized as circular and planar.

Tibiofibular joint kinematics

The loading protocol that was used to elicit tibiofibular motion was described in detail in Asa et al. (2014). Briefly, following surgical protocol the cadaveric specimens were amputated at two fibular lengths (long = 10; short = 5 cm), and in half of the specimens the IOM was sectioned. A suture was applied to the tendon of the BF muscle, the free end of which was attached to the actuator of a Material Testing Machine (Instron®, 8872, Canton, MA, USA; Fig.2). The Instron® was programmed to generate a displacement-controlled cyclical (0.5 Hz) tension with an amplitude of 14 mm. Optoelectronic markers (Optotrack Certus, Northern Digital, Waterloo, Ontario, Canada; Fig.2) were rigidly attached to the tibia and fibula, and using the joint coordinate systems the three-dimension kinematics (flexion/extension, abduction/adduction, rotation) and absolute displacement were calculated.

Figure 2.

Experimental set-up showing the position of the specimen within the materials testing machine.

Data analysis and statistics

To determine the reliability of the surface area and inclination angle measurements, two-way random, intraclass correlation coefficients (ICCs) were calculated (Shrout & Fleiss, 1979), and the following ICC intervals were used to define the magnitude of reliability (Weir, 2005): ICC < 0.4 = poor; 0.4 < ICC < 0.59 = fair; 0.60 < ICC < 0.74 = good; ICC > 0.74 = excellent. If, following reliability analysis, the ICCs were excellent, the mean values were used in a paired one-tailed t-test to evaluate if there was a significant difference between tibial and fibular inclination or tibial and fibular surface areas. To assess the strength of the relationship between the inclination angle and surface area independent of joint shape, Pearson correlation analyses were performed for both the tibia and fibula separately. Chi-square tests were performed to determine if concavity was statistically related to the shape of the articular surface on the fibula and tibia. Finally, multivariate linear regression analyses were performed, using a forward stepwise approach in order to determine the anatomical variables that best predict the angle of abduction and fibular displacement. To ensure that variance was not being artificially inflated, the multi-collinearity of each variable included in the model was assessed using the tolerance value and variance inflation factor (VIF), where values < 0.1 and > 10, respectively, suggest that multi-collinearity may be biasing the regression model (Field, 2005). All statistical analyses were performed using IBM SPSS Statistics (Version 21 Armonk, New York, USA), and statistical significance was accepted at P < 0.05.

Results

The ICCs for the anatomical measurements ranged from 0.88 (tibia inclination) to 0.99 (IOM length), suggesting excellent reliability across all measures (Tables1 and 2). The inclination angle was significantly greater (P < 0.001) on the fibula compared with the tibia (Fig.3a), and the two measurements were moderately correlated with each other (Table1; r = 0.492; P = 0.001). Conversely, the tibia and fibula articular surface areas were not significantly different (P = 0.328; Table1), and the two surface areas were stron-gly correlated (r = 0.80; P < 0.001; Table1). No significant relationships were found between the inclination angle or surface area for either the fibula (r = 0.032; P = 0.851) or tibia (r = 0.130, P = 0.422; Figs3 and 4). Mean (SD) IOM surface areas of 487.3 (169.5) mm² and 1616.0 (197.7) mm² were found for the short and long specimens, respectively, and the difference between them was statistically significant.

Table 1.

Qualitative and quantitative assessment of the tibia and fibula joint articular surface anatomies, the relationship between them and the reliability of the measurements

	Surface area (mm2)			Inclination (°)			Shape (% total)				Concavity (% total)
	Mean (SD)	r	ICC	Mean (SD)	r	ICC	Elliptical	Triangular	Circular	Square	Concave	Trochoid	Saddle	Planar	Convex	Double-trochoid
Fibula	180.0 (57.2)	0.79	0.96	23.1 (7.9)	0.49	0.91	38.5	48.7	12.8	–	43.6	–	15.4	30.8	2.6	7.7
Tibia	193.2 (61.8)		0.98	33.5 (8.9)		0.88	30.8	2.6	59	7.7	7.7	–	5.1	56.4	30.8	–

ICC, intraclass correlation coefficient.

Table 2.

Multivariate linear regression models describing fibular abduction and displacement, including the model diagnostics

						Collinearity
Model	Adjusted R2	Significance	Variables	Beta	P-value	Tolerance	VIF
Abduction	0.411	0.008	Fibular concavity
			Concave vs. saddle	6.6	0.013	0.30	3.31
			Concave vs. planar	3.4	0.071	0.34	2.93
			Concave vs. convex	6.5	0.118	0.59	1.70
			Concave vs. double-trochoid	3.8	0.183	0.44	2.25
			Fibular shape
			Elliptical vs. triangular	5.6	0.001	0.40	2.51
			Elliptical vs. circular	5.6	0.012	0.51	1.98
			Tibial concavity
			Planar vs. convex	3.8	0.047	0.35	2.89
			Planar vs. concave	6.1	0.047	0.40	2.48
			Planar vs. saddle	−4.1	0.281	0.36	2.82
			Tibial shape
			Circular vs. elliptical	−1.4	0.379	0.46	2.15
			Circular vs. triangular	−8.9	0.076	0.41	2.42
			Circular vs. square	−7.2	0.006	0.58	1.73
			Fibula surface area	−0.04	0.001	0.57	1.75
Displacement	0.484	< 0.001	Fibula concavity
			Concave vs. saddle	−0.28	0.905	0.83	1.21
			Concave vs. planar	0.42	0.821	0.82	1.22
			Concave vs. convex	−5.5	0.306	0.86	1.16
			Concave vs. double-trochoid	10.6	0.002	0.90	1.11
			Amputation length	−6.3	0.000	0.91	1.09
			Fibular inclination	0.28	0.008	0.81	1.24

VIF, variance inflation factor.

Figure 3.

Comparison of the mean (SD) inclination angle between the tibia and fibula (a), and the relationship between the two (b) (*P < 0.05).

Figure 4.

Comparison of the mean (SD) surface area between the tibia and fibula (a), and the relationship between the two (b).

With respect to the classification of articular surface shapes and concavity, the fibula were primarily triangular (48.8%) and concave (43.6%), while the tibia were mainly circular (59.0%) and planar (56.4%; Table1). Furthermore, a significant association was found between shape and concavity for both fibular (P = 0.030) and tibial surfaces (P = 0.002; Fig.5).

Figure 5.

Relationship between the inclination angle and surface area for the fibula (a) and tibia (a).

The multivariate linear regression model suggests that the concavity and shape of both the tibial and fibular articular surfaces, in combination with the fibular articular surface, significantly contribute to fibular abduction (Table2). With respect to the fibular articular surface, a joint that is saddle in concavity and either triangular or circular in shape will result in a 6 ° increase in abduction angle. In addition, every mm² increase in the surface area of the fibula will contribute to a 0.04 ° decrease in abduction (Table2). The model that best predicted the magnitude of displacement contained fibular concavity, amputation length and the inclination angle of the fibular joint surface. When the joint displays a concavity described as double-trochoid, the displacement will increase by approximately 11 mm. A short amputation will result in approximately 6 mm less displacement, while the fibula will displace an extra 0.28 mm for every 1 ° increase in the angle of fibular inclination (Table2).

Discussion

To date, this is the first investigation to quantitatively assess the relationship of the anatomical characteristics between the tibial and fibular components of the PTFJ. Overall, there was no relationship found between the inclination angles of the tibia and fibula, but a strong relationship between the surface areas was indicated. Furthermore, this study also revealed the anatomical variables that most strongly predict the magnitude of fibular abduction (fibular concavity, fibular shape, tibial concavity, tibial shape and fibular surface area) and fibular displacement (fibular concavity, amputation length and fibular inclination).

Although Ogden (1974a) presents a PTFJ classification system based on the inclination of the fibular joint surface, it was not used in the current investigation as it does not account for the inclination of the tibia and it is, by admission, an imprecise system. Therefore, in the current work both the inclination of the tibia and fibula were measured, and while no relationship was found between them, a more thorough description of the joint is provided. The difference in tibia–fibula inclination angles may be attributed to the concavity of the articular surface. The distribution of concavities for the tibia and fibula were not similar, indicating that tibial and fibular surfaces do not articulate with mirrored concavity and, when joints articulate with different concavities, it seems reasonable that the angle of inclinations would also be different.

With respect to articular surface areas, Eichenblat & Nathan (1983) reported that fibular articular surface areas were greater than tibial surface areas, while Espregueira-Mendes & da Silva (2006) measured greater tibial compared with fibular surface areas. However, more than half of the samples that were measured in the Eichenblat & Nathan (1983) study were from ancient archeological excavations, and may bring into question the generalizability to a more modern population like the cadaveric specimens used in Espregueira-Mendes & da Silva (2006) as well as in the current investigation. The results presented here, which agree well with Espregueira-Mendes & da Silva (2006), also used a similar surface area measurement technique, different to that of Eichenblat & Nathan (1983).

Recently, Asa et al. (2014) suggested that fibular abduction is affected by the viability of the IOM; however, they did not investigate the anatomical characteristics of the PTFJ. As it is widely known that different types of joints across anatomical locations allow for different types and magnitudes of motion (Barnett & Napier, 1952; Iwaki et al. 2000), it seems reasonable that differences in concavity and shape at the joint surface would also result in joint motion differences (Iwaki et al. 2000). This premise supports the current findings presented here, where the shape and concavity of the appositional joint surfaces affected both the abduction and displacement of the fibula. Furthermore, surface area increases on the fibular aspect of the PTFJ resulted in decreases in the degree of abduction. Although speculative, it is likely that this resulted from the increased friction generated between the joint surfaces. However, it is also important to consider that with an increased surface area comes an increase in the capsular area, which may lead to greater joint stiffness accompanied by decreased motion. With respect to fibular displacement, the length of the amputation was found to be a significant predictor of movement, with longer fibulae abducting to a lesser degree, agreeing well with past work (Asa et al. 2014) suggesting a supportive relationship of the IOM. Further investigation is required to determine whether this relationship is simply an indication that a shorter amputation inherently results in a greater proportion of damage to the IOM (i.e. increased damage to the IOM fibers).

Although the anatomical measurements of the joint were done with great care, and they are highly repeatable, the accuracy of these measurements may have been impro-ved through the use of sophisticated imaging techniques (i.e. computed tomography or magnetic resonance imaging). Furthermore, developing and incorporating image-based measurements would provide the tools required for screening of the PTFJ in actual patients with BKA to screen for those who may be at risk of pathological fibular movement. Another limitation of this work is the use of embalmed cadaveric specimens. Although the embalming process may change the mechanical properties of the tissue, all specimens were prepared with the same solution and thus the effect that this has on the resulting kinematics is controlled for. Finally, although good reliability was found for the anatomical measurements, these were only performed by a single measurer. Therefore, slight increases in the accuracy of these measurements may have been possible with multiple investigators.

Conclusion

The current study presents evidence suggesting that the anatomy of the PTFJ can have significant implications in pathological motions specific to BKAs. Predicting BKA fibular kinematics in vivo may allow the rehabilitation team to preventively treat the patient, reducing the patient's pain, rehabilitation time, while increasing their ambulation and ultimately their quality of life.

Conflicts of interest

The authors have no conflicts of interest to declare.