Clinical and biometrical 12-month follow-up in patients after reconstruction of the sural nerve biopsy defect by the collagen-based nerve guide Neuromaix

Peripheral nerve injury (PNI) can have devastating consequences for patients when reconstructive strategies prove unsatisfactory [1]. Patients not only suffer from the functional motor, autonomic, or sensory loss caused by the disconnection of the nerve, they often develop secondary complications of which chronic neuropathic pain is the most disabling [2, 3].

In clinical practice the autologous nerve transplantation (ANT) is still regarded as the clinical gold standard to repair complex and extensive nerve injuries [4]. Yet the amount of autologous nerve material suitable for transplantation is limited in the human body. To reconstruct, most commonly the sensory sural nerve (SN) is harvested and implanted to bridge the nerve gap caused by the trauma [4]. Moreover harvesting the SN unquestionably leads to functional loss at the donor site, frequently followed by secondary complications, such as wound healing problems and chronic neuropathic pain [3]. Frequently, the autologous nerve material harvested is not sufficient to cover the complete diameter of the damaged nerve. Thus even after reconstruction by ANT, functional recovery is certainly not guaranteed in each case [1].

It is therefore not surprising, that in the last decades many new strategies and a large number of alternatives for the reconstruction of PNI have been explored for their effectiveness in supporting nerve regeneration. For example, other autologous materials [5‐7] (e.g., veins-, muscle-, fat-grafts, and processed nerves from human corpses), non-degradable [8] (e.g., silicon, or poly-glycolic acid), and biodegradable or bio-derived materials (e.g., collagen, poly-lactide, chitosan) with micro-guidance structures have been tested pre-clinically for their merit in supporting nerve regeneration, most commonly in rodent models of PNI [9‐18]. However only a few of these materials were actually clinically evaluated and approved for human use [19, 20]. From design perspective, these materials exist as hollow nerve tubes connecting both nerve stumps by an empty tube lumen. These materials were most often used for bridging small defects in sensory nerves (i.e., most commonly in the hand), where they proved to be effective in leading to some degree of sensitivity. However the bridging of larger nerve defects has remained a challenge. It was suggested that for effective bridging, additional microstructures within the tube lumen were required [21].

Currently, new strategies aim at developing such microstructures using various scaffold designs. For example, it was claimed that human donor nerve allografts retain their connective tissue microstructures and are already clinically used to bridge nerve gaps varying of 5–50 mm in sensory, motor, and combined nerves [22, 23]. Twenty-nine out of 42 acute repairs were performed in sensory nerves. From these acute repairs, the response rate, which was defined as any improvement in quantitative and/or qualitative data investigated, was 92.9%. Moreover similar response rates for the repair of sensory branches of the facial nerve have been reported [24]. So far this allograft is the only clinically used scaffold to our knowledge, providing microstructures to regenerate nerves.

The difficulties of translating animal data into the clinical setting have been emphasized lately [25]. Therefore, we introduced the medial SN biopsy model to investigate newly developed nerve guides in human PNI [20, 25]. In the standard procedure, the nerve gap that arises from nerve biopsy cannot be reconnected without tension. Reconstruction by ANT would be pointless in this case, as sensory loss at the donor site would be the trade-off [33]. When such an overcritical nerve gap persists, this often results in a permanent sensory loss at the lateral aspect of the foot. SN biopsies thus provide a very standardized human nerve injury model to test the biocompatibility and regeneration supporting potential of newly developed nerve guides [25]. Therefore in this current study we implanted the collagen-based, micro-structured nerve guide “Neuromaix” into 20–40 mm nerve gaps caused by medial SN biopsy in patients. Patients were followed until 12 months after surgery. Safety as well as performance parameters (i.e., sensory testing of different sensory modalities) were quantified for every single follow-up visit.

Twenty-three patients were screened according to the in- and exclusion criteria listed in Table 1. Three patients did not fulfill the criteria for implantation. Finally Neuromaix was implanted in a total of 20 patients after medial SN biopsy. These patients consisted of 6 female and 14 male individuals, with a mean age of 52 years ± 1.9 and ranging from 39 to 69 years. In twelve patients additional muscle biopsies (gastrocnemius, n = 10 and/or vastus lateralis, n = 2) were simultaneously performed. During the course of the study, one non-study-related severe adverse event (SAE) and three non-study-related adverse events (AE) were reported. No procedure-related, AE or SAE occurred. Two patients were lost in follow-up (Patient 003 and 010, marked with * in Table 3). Hence, for the evaluation of performance 18 datasets were included in the analysis; however, safety was evaluated for all 20 patients.

The first follow-up visit occurred at 1 month after operation (MPO). All patients investigated in this current study demonstrated clear wound healing during the complete follow-up period. Figure 1 demonstrates representative images at one, three, and 6 months after surgery. Patients did not report foreign body sensations caused by the implanted material. In addition, except for 2–3 days post-operative pain around the area of incision, they were pain-free within 1 week after surgery.

Prerequisite for including the patient’s dataset into the performance analysis was to rule out that the patient’s underlying disease may not allow formation of new sprouts. Hence, neuropathological examination of the biopsied nerve samples was used to distinguish whether newly formed axonal sprouts were present in the nerve tissue. All patients demonstrated clusters of regenerating nerve fibers in the histological preparations, indicative of an intrinsic ability of the nerve fibers to form new sprouts that may regenerate (Additional file 2: Figure S2).

Representative examples of the wandering HTS in time are shown in Fig. 2a–c. Nine months after surgery, the HTS was detected as a sensitive, electrifying spot on the lower leg. Three months later this spot could be detected at the level of the lateral malleolus.

A positive HTS below the complete operation area was first detectable at 6 MPO in 2/18 patients (11%). Thereafter the percentage of patients with a positive HTS at the lower leg increased to 10/18 (56% at 9 MPO) and 15/18 (83% at 12 MPO). From these 15 patients with a positive HTS below the operation area, 11 showed simultaneously a reduction in the ASENS. Five of these 11 reported a HTS located below the lateral malleolus (28%, data points above the dotted 100% line in Fig. 2d). Immediately after surgery and at 1 MPO, the majority of patients reported complete numbness at the ASENS. Thereafter some degree of (protective) touch sensation developed over time, but often patients described a different sensation when compared to the non-operated side (i.e., most commonly described as: “delayed sensation” and/or “sensation as being covered under an asensitive layer”). ASENS did change in time in individual patients, but mean values during the 12-month follow-up did not statistically significantly differ from the status immediately after surgery (Fig. 2e). Two patients who reported a clear HTS at 9 MPO did not sense a positive HTS at 12 MPO anymore, which coincided with the recurrence of full sensation within the ASENS.

Figure 3 demonstrates in a–c examples of the ASENS in relation to the location of the positive HTS 1 month after surgery, whereas a_*–c_* shows a smaller ASENS at 12 MPO. In addition, Fig. 3a**, b**, c** visualizes the distance traversed by the HTS in relation to the ASENS at 12 MPO (Red circles show the location of the HTS and yellow arrows the location of the middle of the scar, respectively).

Prior to operation, mechanical sensation at both feet was similar (operated, open circles: 3.8 ± 0.3, non-operated, filled squares: 3.8 ± 0.2, Fig. 4a). After the operation, a clear elevation of the mechanical sensitivity threshold could be observed at the operated foot when compared to the non-operated side (operated: 5.8 ± 0.4, non-operated: 3.8 ± 0.2, ***p < 0.001). Immediately after surgery, thresholds were in the range of “loss of protective sensation.” At this time point, patients often responded with “no sensation at all” (53%, deep pressure sensation only). This post-operative elevation of the mechanical sensitivity threshold faded in time, but thresholds remained higher than compared to the non-operated side (Fig. 4a, **p < 0.01 at 3 and 9 MPO and at 12 MPO; operated: 4.5 ± 0.3 open circles, non-operated 3.5 ± 0.2, filled squares *p < 0.05). This coincided with patients reporting return of protective sensation (i.e., complete numbness changed into some degree of touch sensation).

Patients displayed post-operative pain around the operation site during the first 2 days after surgery, but after 1 week, patients were completely free of any pain. Over time patients did not show strong fluctuations in pain sensation at the lateral foot (Fig. 4b). VAS scores prior to surgery were slightly elevated for the operated side (5.2 ± 0.7, open circles) when compared to the non-operated side (4.0 ± 0.8, *p > 0.05, filled squares). None of the patients reported pain, but rather reported numbness, or tingling sensations after touch of the ASENS. Immediately after surgery, a drop in vibration sensation (Fig. 4c) could be observed at the operated side compared to the preoperative status, but this reduction was not statistically significant (preoperative: 5.1 ± 0.8 and 0 MPO: 2.4 ± 0.9). Vibration sensation returned (12 MPO: operated: 5.0 ± 0.6, open circles), but remained slightly lower than at the non-operated side (12 MPO: 5.7 ± 0.4, filled squares).

Figure 4d demonstrates the percentage of patients that perceived blunt stimuli (10 gr SWT filament) in time. Blunt sensation returned quite rapidly after surgery, as with 1 MPO approximately halve of the patients (56%) sensed the 10 g SWT filament. After 12 MPO, this percentage was increased to 83%. Sharp sensation (pin-prick) did not recover as quickly as blunt sensation after surgery; up to 9 MPO less than half of the patients were able to detect the sharp stimulus (33%). During the last 3 months, a clear recovery could be observed, as finally at 12 MPO 61% of the patients were able to detect sharp stimuli (Fig. 4e). The number of patients who detected the cold stimulus increased only little in time. Even after 12 MPO the majority of patients were unable to detect cold stimuli (33% Fig. 4f).

To date there is still an urgent clinical request for alternatives to the autologous nerve transplantation for the repair of damaged peripheral nerves [1‐3]. These alternatives can exist of biomaterials having the advantage that they can be used off the shelf. In this present study we evaluated the newly developed collagen-based nerve guide Neuromaix in a first in human study. As a two-component scaffold, Neuromaix was implanted into 20–40 mm sural nerve biopsy defects to support functional axonal regeneration across this nerve gap [25]. Safety as well as performance parameters of this application of the scaffold were evaluated up to 1 year after the implantation.

Our data first of all show that Neuromaix is safe; uneventful wound healing was evident in every patient during the complete follow-up of this study. Our findings are in line with previous reports about collagen materials being well tolerated by the human body, in general [36] and more specifically after implantation as a tube in human PNI [37]. Moreover, our own findings for Neuromaix in animal experiments of PNI complement these current first in human results [12‐14, 26‐31]. In addition, patients did not complain about foreign body sensations caused by the material. Except for a few days post-operative pain (maximally 1 week), patients did not develop any persistent pain around the scar area or at the lateral foot. Most commonly sensation after touch at the ASENS was described as “delayed” or “covered by an asensitive layer” hypoesthesia, but never as painful to a non-painful stimulus (mechanical allodynia), or as burning or stinging pain (hyperalgesia).

Second, our data support the notion that implantation of Neuromaix may serve as a standard repair method to overcome secondary complications after nerve biopsy [37‐41]. Biopsy removal at the level of the calf, in comparison to the classical method at the lateral malleolus [39], provides, as our data confirm, ideal wound healing conditions with excellent soft tissue covering. This method provides the opportunity to obtain muscle biopsies at the same time, without losing diagnostic value of the biopsies taken [42, 43] and providing the most optimal wound healing conditions for the patient, as already discussed previously [25]. In addition, the majority of the patients reported spontaneous tingling and electrifying sensations, which has been associated with movement of the growth cone, but no pain between follow-up visits [32, 33].

The nerve’s intrinsic ability to regenerate is, of course, a prerequisite for the evaluation of performance. Patients assigned for a nerve biopsy concern a population that has already been examined clinically by the standard diagnostic work-up for PNP, showing no abnormalities in blood/spinal fluid counts, though often demonstrating abnormal electrophysiological parameters (i.e., impaired amplitudes and motor/sensory nerve conductance velocities). By semi-thin resin section histology of the biopsy material, we confirmed the presence of regeneration clusters in every patient, indicating the intrinsic capability of the nerve to regenerate. Thus, all patients with complete follow-up could be included into the performance analysis (n = 18). However, we noticed by tracking the HTS in time that regeneration rates in our patients were slower than the 1 mm/day frequently reported in literature for healthy persons [44]. Our data confirm that the patient’s underlying disease condition is an important factor influencing nerve regeneration [45]. Therefore the patient’s underlying condition should be taken into account all time while planning and evaluating data in this nerve injury model, as it has an important influence on regeneration and the timing of the return of sensation.

Moreover another important factor in our study to consider is the age of our patient population (mean age 52 years ± 1.9, range 39–69 years). Previous work from others already showed that age is an important factor influencing myelin- and axonal-related parameters, the proportions of different fiber types in the nerve, and electrophysiological properties (Summarized in Verdu 2000 [45]). It can be expected that variability in age may also account for the individual differences observed in our patients with regard to the timing and degree of returning of sensation.

Indeed biopsy and implantation at the level of the calf is located more distant from the target organ than at the ankle [37‐41], and therefore regenerating nerve fibers have to traverse a longer distance to re-innervate the lateral skin of the foot. It can be expected that this has also implications for the timing of return of sensation. However, we were able to detect positive HTS in fifteen out of eighteen patients by 12 MPO. Moreover eleven of them demonstrated a HTS below the operation area with simultaneous reduction in the ASENS, suggesting that nerve fibers once regenerated across Neuromaix, entered the former trajectories of the SN [32, 33]. Five of the latter demonstrated positive HTS that passed the lateral malleolus heading towards the lateral aspect of the foot and two of them reported complete return of sensation by 12 MPO.

All patients reported reduced sensation in the area innervated by the SN immediately after operation. This was confirmed by the testing of various sensory modalities. Some degree of protective touch sensation returned in all patients. Sharp and cold sensation recovered more slowly than touch and vibration sensation. As previously mentioned, timing of the return of sensation has been shown to depend on age, gender, underlying disease of the patient, time to repair and the type, location, and extend of injury [46]. This may partly account for the individual differences in recovery seen in our population (i.e., here influenced by age, gender, and underlying disease). With respect to the different sensory modalities, thresholds generally increase during aging [45]. Comparison to other repair approaches is difficult, as the timing of, and the extent of sensory regain highly depends on the size of the nerve gap to be bridged (i.e., < 15 mm or overcritical), the distance to be traversed by the regenerating axons (i.e., distant or near the target skin areal) as well as the location of the area to be re-innervated (i.e., hand palms/foot soles, or dorsolateral foot). For the lower extremities, the greatest alterations in sensation after PNI have been reported to manifest as loss of mechanical pressure, vibration, and cold detection [47]. Hyperesthesia as well as hyperalgesia were less frequently seen. In addition pain-related symptoms were detected with reduced thresholds to cold and pressure pain. In contrast, warm pain sensation showed elevated thresholds [47]. Loss of touch, vibration, and cold sensation was also evident in our patient population after surgery, but steadily improved over time.

Eleven of fifteen patients with a positive HTS at the lower leg demonstrated simultaneously a reduction in the ASENS, suggesting that newly regenerated fibers re-innervated the skin areal of the lateral foot. As these observations were not concurred by pain-related symptoms at the ASENS, implantation of Neuromaix could become a standard procedure after nerve biopsy to overcome possible co-morbidities. Restoration the transmission of sensory information from the periphery to the brain has been suggested to alleviate pain-related symptoms, such as allodynia [48, 49]. This method can even be extrapolated to conditions where harvesting of donor nerve material for transplantation leaves a nerve gap.

Electrophysiological recordings and neuro-imaging techniques however are needed to complement these first results on sensory regain [50, 51]. Nevertheless, our data provide promising prospective for the reconstruction of combined nerves, which should be tested in larger, randomized, multi-center, clinical trials in the near future.