Introduction

In 1915, Delorme [4] became the first to describe using joint flexion to diminish the gap between two nerve stumps. His technique was popularized from the First World War through the 1930s, but fell into disuse afterwards. Its detractors maintained that, after the immobilization period, joint extension would place tension again at the suture site and secondary rupture would occur [6]. The later emergence of nerve grafts [14] contributed to the low popularity of Delorme’s technique, though it was never abandoned entirely, and recently has experienced increased use [1, 2]. Recently, we described modifying Delorme’s original technique to perform nerve suture under flexion in a pediatric patient, which included monitoring the nerve repair site with serial ultrasounds (US), a specific, intense physiotherapy program, and use of an articulated orthotic device to gradually extend the flexed joint under US control. Our full and complete protocol for progressive joint extension has already been published, if more precise information is required [20].

In the present paper, we describe a series of eight patients who underwent a direct nerve suture under joint flexion. Our objective was to determine how effective our multimodality treatment was at avoiding the use of grafts, and at obtaining faster and more complete functional nerve recovery. A secondary objective was to determine the sensitivity of US or, if required, magnetic resonance imaging (MRI), for identifying nerve dehiscence requiring either revision or replacement of the nerve suture, the second surgery using nerve grafts.

Methods and materials

Data from eight consecutives lacerating nerve sutures performed under joint flexion between January 2016 and December 2018 were collected prospectively. As an open injury followed by a nerve deficit suggests the presence of nerve section, the procedure was indicated as soon as possible.

For nerve gaps of 6 cm or shorter, the option of performing nerve repair under joint flexion was considered intraoperatively, based upon the length of the nerve gap. Before enrolling each patient, the pros and cons of this type of nerve repair were discussed with both the patient and/or, in pediatric cases, the patient’s parents. All ethics committee standards were respected. Though meticulously described in a previous article [20], a brief description of our nerve repair protocol using joint flexion follows.

After exploration, the diagnosis of complete nerve section was made, in all cases, using either direct visualization, nerve stimulation, or nerve action potentials (NAPs). Nerve stumps were trimmed until healthy tissue was observed, and the remaining nerve gap measured. Both nerve endings also were dissected, trying to leave as minimal a gap as possible. Depending on the affected nerve, either the knee or the elbow was flexed to determine if direct suturing was possible. If this was considered feasible, two initial equatorial 7.0 anchorage stiches were placed, while the suture was completed with separated nylon 9.0 stitches. Hip or shoulder mobilization was performed after the nerve repair and before wound closure, to ensure that these movements would not stretch the nerve post-operatively. After determining the minimum degree of joint flexion at which no tension was induced at the suture line, the limb was held immobilized throughout the remainder of surgery and during cast placement.

After a window was created in the cast 3 weeks after surgery, a baseline ultrasound (US) was performed to confirm nerve repair preservation. Two weeks later, the cast was removed and an articulated orthotic device applied, maintaining the joint (knee or elbow) in its initial degree of flexion. From that moment onward, an intense physiotherapy program was initiated to recover joint flexion and prevent joint stiffness; however, extension was not permitted beyond the critical point, and just joint flexion and not-maximal extension were trained. A second ultrasound was obtained 3 weeks later to confirm not only nerve suture preservation but also vascular flow at the repair site and nerve gliding under joint flexion. If all three previously mentioned requirements were confirmed, from that point onward, always under strict US guidance, extension gradually was increased by altering the degrees of orthotic device flexion in increments of 10° every 3 weeks. This procedure (physiotherapy ultrasound- ≤ 10° joint extension guided by ultrasound) was repeated until either (1) complete joint extension with nerve suture preservation was accomplish or (2) repair site rupture was observed, in which case repeat exploration and repair were indicated, this time using nerve grafts to bridge the gap. In all cases, our physiotherapy program was followed to completeness.

Results

Table 1 provides individual patient demographic and clinical data, and summarizes each patient’s clinical course after treatment. No patients were lost to follow-up. One patient presented with an upper limb nerve injury (radial nerve, #5), while the rest (n = 7) had suffered a lower limb injury, either of the entire sciatic nerve (n = 2) or just its peroneal component (n = 5). All patients were male, and ranged in age from 6 to 42 years old (mean = 21.6 years). Nerve gaps deemed amenable to repair ranged from 2.5 to 6.0 cm (mean 4.8 cm) and the range of joint flexion required for tension-free nerve repair ranged from 30° to 90° (mean = 50°).

Table 1 Patients included in the present series and results obtained after nerve suturing under joint flexion
Full size table
Fig. 1
figure 1

Patient #3. A A 12-year-old boy suffered a domestic lacerating injury that completely sectioned his right peroneal nerve at the popliteal fossa, 4 cm distal to the sciatic nerve bifurcation. A foot drop was immediately observed. During surgical exploration, a 4-cm gap was repaired using a direct nerve suture with his knee in 45° flexion, a degree of flexion then maintained with a cast. B An articulated knee orthotic was placed 4 weeks after the surgery, maintaining the same degrees of articular flexion as with the cast. At that time, the nerve suture was evaluated with ultrasound, and was found to be intact. C MRI acquired in coronal plane with reconstruction of the common peroneal nerve showing the proximal (small arrow) and distal (long arrow) pathway, with dehiscence of the neurorraphy and a significant thinning of the common peroneal nerve (arrowhead). D Surgical re-exploration was indicated, and suture rupture detected. E The new gap was 5 cm after trimming the nerve stumps (1 cm longer than during the primary repair). F Sural nerve grafts were employed to bridge the gap. G Two years later, showing M4 recovery of peroneal right function

Full size image
Fig. 2
figure 2

Patient #4. A 31-year-old male suffered a lacerating injury with a knife to the posterior surface of his distal thigh. A A sciatic nerve gap of 6 cm was identified. B The gap was closed with a direct nerve suture with his knee in 90° of flexion. Fifteen months post-operatively, the patient had recovered foot sensation, as well as plantar and dorsal flexion

Full size image
Fig. 3
figure 3

Patient #5. A 6-year-old child suffered a lacerating injury to the left arm and elbow after falling from a horse and hitting a wire fence. A A 4-cm nerve gap in the radial nerve was found after trimming the nerve stumps, and a direct suture placed kee** the elbow flexed at 90° for 2 months in a cast. After several follow-up, ultrasounds showed that the repair was intact; progressive elbow extension was performed under US guidance, supervised by his physiotherapist. B The first signs of nerve recovery were present at 3 months, with complete restoration of radial nerve function documented 6 months after the accident

Full size image

Interestingly, half of our patients (n = 4) sustained with a nerve rupture detected at the second ultrasound that required secondary nerve repair using nerve grafts and therefore avoiding joint flexion. In all four cases, US revealed dehiscence at an early stage (cases #3, 4, 6, and 7, mean time = 5 weeks, range 3–8). In the remaining four patients, in whom the nerve suture was preserved and confirmed by US, an early and very good response (M4 in all cases, starting at 3 to 8 months, mean = 4.25 months) was observed. The adopted imaging method was 100% accurate at identifying nerve suture preservation. Of note, in case #8, peroneal nerve-dependent muscle contraction was observed on US at 2.5 months, 2 weeks before any contraction was observed clinically. The early detection of nerve failure permitted early re-do surgery with grafts (cases #3, 4, and 7) or a tendon transfer (case #6) and, ultimately, obtaining good results in all four patients in whom nerve suture rupture was identified (M4 in two patients, M3 in one, and a tendon transfer in one). Tibialis tendon transfer was indicated in case #6 because the COVID-19 pandemic prevented nerve grafting in a timely enough manner.

The mean follow-up duration was 27.6 months (range = 12–48). In the four patients who experienced suture rupture, either incorrect use of the immobilization device, excessive movement (i.e., accidentally falling), or a broken device was considered the probable cause of dehiscence.

Discussion

This is, to the best of our knowledge, the largest reported series in which modern imaging and adequate physiotherapy were combined to document the results of nerve suturing performed using joint flexion as a way to avoid utilizing nerve grafts to bridge the gap when repairing an injured nerve. Of course, a failure rate of 50% is high, and raises the question: should we keep on using this type of nerve repair?

Suture failure avoidance is perhaps the key-point when using the joint flexion technique for nerve suture. As mentioned in the results, problems related to the misuse of the immobilization device or doing violent movements by the patients should be avoided in order to lower the rate of nerve suture dehiscence to the maximum, or eventually to even zero.

Our series clearly shows the pros and cons that should be considered in the future in order to maintain or abandon this technique.

The first advantage is that, if successful, avoiding nerve grafts generally yields a better result than using a graft, not just in our series but in the literature. The outcomes of nerve grafting are poorer than those of direct sutures, whether the suture is a primary (up to 7 days after trauma) or secondary suture (thereafter) [21]. Obviously, this is because the growing axon needs to transverse just one nerve suture, that being the distal stump of the nerve itself, instead of sutures at both ends of a graft. It has been calculated that the delay required by axons to transverse a suture is approximately 30 days in transected nerves [11, 13, 15]. Hence, direct nerve sutures are associated with quicker arrival of nerve impulses to the target muscle and, consequently, less denervation time and a faster and better final result, relative to classical graft repairs. This is especially important when sensory and motor targets are located at considerable distances from the site of injury, as in sciatic nerve injuries in the buttock or thigh [13]. Also, the regeneration of axons in a mixed nerve into inappropriate pathways is a major contributing factor to poor functional recovery, and this is probably more likely the case with nerve grafts than with direct sutures [3, 11, 13, 15].

Positive results of nerve sutures performed using joint flexion are reported in the literature. Oberlin reported sutures being applied during knee flexion for sciatic nerve repair in a series of 12 patients, with the majority experiencing grade 3 or 4 muscle recovery in flexor, extensor, and fibular muscles [16], while Pfister and Oberlin demonstrated their utility in cadaveric models [18]. Such good results have been replicated by Döring et al. and by Mathieu et al., who have both reported small series of direct sciatic nerve repairs under knee flexion [5, 13]. The favorable results of these series and our own cases are encouraging, compared to the less-satisfactory outcomes reported in previously published series involving nerve grafts [8,9,10,11, 19]. However, since larger series and randomized clinical trials remain elusive, the efficacy and safety of this technique have yet to be proven. In the end, the effects of nerve stretching remain an enigma [12].

A second encouraging finding in our series is that US was so accurate (100% sensitive and specific) for distinguishing nerve suture rupture from preservation, which implies that timely secondary nerve repair with grafts can be assured, providing a good window of time for acceptable recovery of neurological function, even in patients in whom the initial procedure fails (albeit, with outcomes perhaps not quite as good as via direct repair). This was, indeed, observed in the three out of four patients in our series who were able to undergo a secondary surgical procedure in a timely manner, two of whom also ultimately experienced M4-level recovery. The fourth patient with failed direct nerve suture also obtained a good (M4) result with a tendon transfer. Ultrasound is considered the best technique to evaluate the surgical repair of peripheral nerves, particularly with large and superficial nerves [17].

On the other hand, a 50% rate of success with nerve repair is definitely not as high as anyone would like to see. The post-operative care that suturing during joint flexion requires also is much more complex than that generally necessary following classic nerve grafting and requires a highly compliant patient. In fact, all instances of nerve dehiscence in our series were attributed either to a failed orthotic or to inadequate adherence to the prescribed treatment. Recently, another clever—albeit more invasive—way to avoid the problems that we observed in our series has been proposed, entailing bone shortening and using external fixators to slowly overcome bone defects without introducing tension to nerves (Allan Belzberg—personal communication). Other “nerve lengthening “ techniques have recently been described in rats, providing further promising strategies in the mid-term future [7].

Another drawback of the technique we have described is that prolonged immobilization normally produces joint stiffness, which is well tolerated in children and young adults, but might be less suitable for elderly or even middle-aged patients. This said, our oldest patient was 42 years old at the time of surgery and did not experience this problem.

In any case, prior to performing nerve repair under flexion, detailed discussion with the patient and/or the patient’s parent/caregivers is required, explaining both the advantages and disadvantages of this approach and the compliance that is needed both from them and from the therapeutic team, if a nerve suture under flexion technique is to be used. It is quite important that the patient and its family understand the importance. In our opinion, this technique could be used in well-selected patients, as its final results are quite good, probably much better than those achieved with classical graft nerve repair. The higher cost of post-operative care also should be explained [20]. Ultimately, full compliance with all aspects of the post-operative protocol is necessary if nerve rupture at the suture site is to be avoided. If such compliance cannot be ascertained, some other approach may be more prudent.

Conclusions

Our main conclusion is that applying nerve sutures during joint flexion to avoid using grafts has a high rate of nerve dehiscence, but offers unusually good and fast results, in terms of neurological outcomes, when it is successful. If this approach is selected for use, it is mandatory that ultrasound also be employed to closely monitor suture status, together with physiotherapy providing progressive, incremental, US-guided extension of the flexed joint to protect the neurorrhaphy site. If nerve rupture occurs, the close monitoring dictated by this protocol should ensure the timely application of successful graft repair. A high degree of compliance is required, both from patients/parents and from the therapeutic team, if nerve dehiscence rates are to be avoided below what is published herein. Further studies with larger numbers of patients and variations in immobilization techniques could provide more definitive conclusions in the future about this technique’s role in nerve repair surgery.