Introduction

Spinal fusion has been the gold standard until the advent of motion sparing technologies, i.e., the total disc arthroplasty, which were first introduced in the mid 1980s [1–4]. Clinical fusion studies indicate that standard lordotic interbody cages seldom improve the patient’s sagittal balance, but generally, they improve the disc height and fusion rate [5–7]. Disc arthroplasty was introduced as an alternative to fusion and to prevent possible known complications following lumbar fusion. The technology has been found to allow for a more physiological spinal alignment and improve bone–implant interface during the initial loading and healing phase [8–11]. These two factors have been found important for successful clinical outcome and for avoiding late complications (implant subsidence, loss of lordosis correction, and adjacent disc degeneration) [4, 12, 13].

The concept of a two-piece ALIF fusion device is generated from a study indicating that the early stage disc arthroplasty implants, intended to move forever but ‘mistakenly’ ended up fusing, ironically, had superior outcomes in comparison to ‘successful’ non-fusion implants [14]. An hypothesis for these findings could be that the degenerated spine benefits from allowing alignment adjustment, meaning better alignment between implant footprint and vertebral endplates; combined with the fact that lumbar degenerated spines quite often auto-fuse [7, 15].

The aim of the present study is to compare the bone–implant relative motion and rigidity of two conceptual different cage designs, a two-piece ALIF cage versus a traditional one-piece ALIF cage in a 360° setting.

Materials and methods

In a human lumbar cadaver model, we analyzed the segmental relative motion and its impact on the implant–endplate interface, using two different ALIF implant concepts; the two-piece ALIF cage (Statur®-L, FBC Device ApS, Denmark) and the more commonly used one-piece ALIF cage concept (Pezo™-A, Ulrich GmbH & Co. KG, Germany). Both are commercially available in Europe. The Statur®-L cage is CE-marked with the use of supplemental fixation, (e.g., pedicle screws or anterior plate) and therefore, a pedicle screw test model was used to mimic a clinical relevant test scenario. Both implants are made of PEEK, have quite similar foot-prints, the same height and surface texture to lock to the endplate. The two-piece ALIF cage (Statur-L) is capable of continuously adapting from 7° to 21° of lordosis in situ, without external adjustment or insertion of additional components. The two-piece technology allows the optimization of segmental lordosis and the cage system will automatically adapt to the supplemental fixation (screws or plate). The two components are placed on top of each other, without any internal locking system (Fig. 1). After insertion, the implant aligns to the vertebral endplates as soon as the inserter is released from the implant (self-adapting). The implant has a keel to avoid anterior–posterior implant migration.

Fig. 1
figure 1

Design of the two tested ALIF cages. Statur-L has a top and a bottom component with the two elements placed on top of each other. The implant has a stabilizing keel on the bone articulating surface. The Pezo implant is a traditional one-piece ALIF cage

Full size image

In comparison, the more commonly used one-piece ALIF cage is born with one specific lordosis angle (Fig. 1). A pedicle rod-screw system (tangoRS™, ulrich GmbH & Co. KG, Germany) was added to mimic clinically relevant mechanical circumstances (360° fusion) (Fig. 2).

Fig. 2
figure 2

a Illustration of the challenge for a one-piece cage to adapt to different RoM and loadings and thereby keeping a good bone–implant interface. b Illustration of the optimized bone–implant interface found using a two-piece implant. The study documents that a significant RoM takes place also in a 360 fusion model

Full size image

Seven fresh-frozen human lumbosacral spines (L3-S1) with a median age 55 years (50–56 years) were used for the study. They were obtained from a tissue bank (Anatomy Gifts Registry, Hanover, MC, USA). The study was approved by the ethics committee board of the University of Ulm (Nr. 383/12). The specimens were kept frozen at −20 °C and thawed overnight at +6 °C before testing. To exclude spinal pathologies, latero-lateral and antero-posterior X-ray films were taken prior to testing. All surrounding soft tissue was dissected, while care was taken to preserve the biomechanically relevant structures. The proximal (L3) and distal (S1) ends of the specimens were embedded in polymethylmethacrylate (PMMA) cement (Technovit 3040, Heraeus Kulzer, Wehrheim, Germany). Special care was taken to align the L4–L5 intervertebral disc horizontally. To mount the specimens in the testing apparatus, flanges were fixed to the PMMA blocks. To prevent desiccation, specimens were sprayed with 0.9% NaCl physiologic solution during testing.

Flexibility testing was performed in a universal spine tester [16] at room temperature. A gimbal fixed to the cranial end of the specimens allows the application of pure bending moments in the three principal motion directions: lateral bending right/left, flexion/extension and axial rotation left/right. During loading, the specimens were allowed to move unconstrained in the remaining five degrees of freedom. The specimens were loaded with pure moments of ±7.5 Nm in all principal motion directions at a loading rate of 1 °/s (except 0.5 °/s in axial rotation). The resulting three-dimensional intervertebral motion was captured with an optoelectronic motion analysis system (Vicon MX 13, Vicon, Oxford, UK) (Fig. 3). Following international standards [17], range of motion (RoM) and neutral zone (NZ) were calculated from motion data.

Fig. 3
figure 3

Specimen equipped with retroreflective markers mounted in the spine tester (center). Six infrared cameras are arranged circumferentially to capture the three-dimensional motion of the spine

Full size image

To measure the relative motion between the interbody fusion implants and the adjacent endplates during flexion/extension, a fluoroscope (Exposcop CB7-D, Ziehm, Nuremberg, Germany) was connected to the spine tester (Fig. 4a). Adapter plates ensured a reproducible positioning of the fluoroscope. Using this approach, the image plane was always exactly perpendicular to the primary motion plane. Video films were taken during the third full motion cycle, and the frames showing the specimen in maximal flexion and extension (turning points) were isolated and used for further analysis. Relative motion was calculated from the fluoroscope images in maximal deflection (Fig. 4b, c) using Photoshop (CS4, Adobe Systems, San Jose, CA, USA). Continuous lines connect the radiographic markers of the interbody implants, while dotted lines represent the endplates. Angles between the upper/lower dotted line (endplate) and upper/lower continuous line (implant) of the respective images were calculated (2 angles at each pair of maximal deflections). The difference in each angle (upper/lower) between the two maximal positions was calculated, and the two differences summed. This sum represents the relative motion between implant and endplate. Care was taken to exactly reproduce the alignment of the lines between the images of the same specimen in maximal deflections, while inaccuracies in the absolute positioning of the anatomical landmark lines were of minor importance due to the calculation of relative changes.

Fig. 4
figure 4

a Fluoroscope affixed to the spine tester; b fluoroscope image of the segment instrumented with the Statur®-L implant in maximal flexion; c fluoroscope image of the treated segment instrumented with the same implant in maximal extension. Dotted lines are drawn parallel to the endplates, while continuous lines connect the X-ray markers of the implant. Since this cage consists of two components, it can follow the movement of the vertebral endplates

Full size image

All specimens were tested in the following three conditions (with alternation of 2 and 3):

  1. 1.

    Intact specimen.

  2. 2.

    One-piece ALIF cage (Pezo™-A, Ulrich GmbH & Co. KG, Germany) with pedicle rod-screw system at L4–L5 (tangoRS™, Ulrich GmbH & Co. KG, Germany) (One-Piece).

  3. 3.

    Two-piece ALIF cage (Statur®-L, FBC Device ApS, Denmark) with pedicle rod-screw system at L4-L5 (tangoRS™, ulrich GmbH & Co. KG, Germany) (Two-piece).

The intact condition served as control for all subsequent steps.

The implantations were carried out according to the manufacturers’ guidelines by an experienced surgeon. Adequate implant size was chosen individually to ensure good performance of the implants.

For the anterior fusion procedures, the intervertebral disc space was prepared with a subtotal anterior discectomy, which consisted of removal of the entire nucleus pulposus and cartilaginous endplates with pituitary rongeurs and curettes, leaving the lateral and posterior aspect of the anulus fibrosus intact. The ALIF cages were chosen slightly higher than the disc space following the distraction–compression principle. The Pezo™-A was chosen as benchmark because it provides a foot-print comparable to the Statur®-L device. This allows testing the implants in an alternating order without compromising performance. Pedicle screw entry points were prepared with an awl. The self-tapping screws were inserted without pilot holes and rigidly connected with titanium rods and locking screws.

Data were analyzed using statistics software (SPSS 19, SPSS Inc., Chicago, IL, USA). After testing for normal distribution (Shapiro–Wilk test), a one-way analysis of variance (ANOVA) with Tukey post hoc test for intervertebral motion data (RoM) and a two-sided Student’s t test for relative motion data was performed to check for significant differences. The significance level was set to α = 0.05.

Results

Significantly less motion at the implant–endplate interface was found for the two-piece device (1.0° ± 0.6°) in comparison to the one-piece ALIF cage (4.2° ± 1.7°) in flexion/extension (Fig. 5). At the same time, no statistically significant differences were found in the overall segmental RoM between the one-piece or two-piece ALIF cage in the 360° set-up, i.e., 3.5° ± 1.9° for the one-piece and 4.2° ± 1.8° for the two-piece ALIF in flexion/extension. Both configurations significantly reduced the RoM in flexion/extension, lateral bending and axial rotation compared to the intact condition in a 360° setup (p < 0.05) (Fig. 6).

Fig. 5
figure 5

The relative motion between implant and endplate of the treated segment for the one-piece and the two-piece implant in flexion/extension

Full size image
Fig. 6
figure 6

Range of motion and neutral zone (gray bars) of the two implants in comparison to the intact state in the three principal motion planes

Full size image

Discussion

The two-piece ALIF cage significantly reduced the relative motion at the bone–implant interface without compromising stability. This means that with this two-piece self-adapting cage technology, an unchanged mechanical bone–implant interface could be measured in the total range of flexion–extension. In comparison, the one-piece cage lost anterior cage–endplate contact when the spine moved into extension. The one-piece cage consequently ends up having a total implant load bearing at the posterior part of the implant which could result in a “stone-in-the-shoe” phenomenon [14]. This is an expected scenario when a patient stands and the spine is relatively extended.

Expandable cage technology is a recent innovation and an increasingly popular alternative to one-piece cages. Some of the implants are relatively complex in their design, and complication data are still limited [18–20]. The biomechanical behavior of the tested two-piece, self-adapting ALIF cage appears to be different than multi-piece cages that expand, though the measurements used differs between the studies. As the height of the multi-piece expandable cages is increased, the lordotic angulation of the expandable implants has an impact on the contact force and the location of the center of pressure and can induce significant stress concentration at the endplates [18]. In contrast, the two-piece, self-adapting ALIF cage does not change in overall height. Therefore, we would not expect a stress-concentration effect at the endplates. The lack of relative motion between implant and endplate found in the current study supports this statement. However, direct measurement of pressure under both types of cages would be required to confirm this.

The biomechanical bone–implant interface, has been shown to be of importance for the load sharing and stress distribution, which again is stated to have an impact on subsidence and thereby loss of the intended correction (segmental alignment and height) [21]. When taking the above into account, it is important to note that the two-piece device did not increase the overall segmental flexibility when used with supplemental instrumentation. The one-piece implant and two-piece implants did not show significantly different segmental movement; but in the two-piece group, the minor movement is taking place within the implant pieces and in the one-piece group between endplate and implant. The current study has some limitations common for biomechanical in vitro experiments. First, the results represent only the immediate postoperative stability of the various implant configurations and do not include the effects of tissue remodeling or long-term viscoelastic response. It is important to keep in mind that cage subsidence is a multifactorial complication involving endplate preparation, strength of additional instrumentation (screws or plates), choice of bone graft material, bone quality etc. Additionally, the physiological loads in the lumbar spine are still not completely known. It was therefore decided to apply pure bending moments, so that the load applied to the specimen was constant along the length of the specimen. This loading scheme was chosen according to the recommendations of Wilke et al. [17] and is thought to approximate estimated daily in vivo peak torques [22], but is low enough not to damage the cadaveric motion segment.

Conclusion

In comparison to the traditional one-piece ALIF cages, the two-piece ALIF cage concept reduced the relative motion at the bone–implant interface without compromising stability. Clinical studies are needed before concluding on the potential benefits related to this new two-piece self-adapting fusion technology.