Functional evaluation of early tendon transfer for thumb opposition in median nerve palsy

Abstract

Background

There is paucity of literature on early tendon transfer in surgical rehabilitation of hands with median nerve injuries. Since the single most important functional deficit in median nerve palsy is the loss of thumb opposition, we evaluated the results of early tendon transfer in restoration of thumb opposition.

Methods

This prospective study involved 10 cases of isolated median nerve paralysis (axonotmesis or neurotmesis) that underwent early tendon transfer for restoration of thumb opposition. A pre- and post-operative evaluation in terms of power and precision grip strength and range of opposition of thumb was done. Median nerve exploration was performed in 4 cases and was supplemented by a tendon transfer for thumb opposition. The extensor indicis proprius (EIP) opponensplasty was performed in nine out of ten cases. In one case where the patient had scarring over the EIP tendon, palmaris longus (PL) opponensplasty was performed.

Results

The median age at injury was 29 years (range; 8 years–57 years). Minimum period of follow-up was 6 months. Six patients who underwent EIP transfer had excellent opposition while 3 had good opposition. The patient, in whom PL opponensplasty was done, had an excellent opposition. The median time for return to work was 2.75 months.

There was significant improvement in the power grip and all three types of precision grip at 6 months follow-up. The percentage deficit in the affected hand compared to the normal hand was significantly reduced. There was no case of tendon pull out in our study, nor did any of our patients have an extensor lag following EIP transfer.

Conclusion

Early tendon transfer has a unique role in the management of median nerve palsy hand and we suggest this procedure should be considered in both high and low lesions.

Type of study

Prospective observational case-series.

Level of evidence

Level IV.

1. Introduction

Low median nerve palsy primarily causes the loss of thumb opposition, and sensations over the lateral three and a half digits. The paralysis of the two radial lumbrical muscles may be of little consequence if the ulnar nerve is intact. The important additional functional deficits caused by high median nerve palsy are the loss of forearm pronation, wrist flexion and radial deviation, index finger flexion, and thumb flexion.¹

The hand function depends, to a large extent, upon a mobile opposable thumb and the radial two fingers. Hence, attempts should be made to restore a mobile opposable thumb as early as possible. Delay may result in joint contractures, making full recovery highly unlikely especially in high nerve lesions where nerve regeneration to affected muscles is likely to take months. Burkhalter, one of the fore most proponents of early tendon transfer, felt that an early tendon transfer works as a substitute during the regrowth of the nerve, thereby improving the function and frequently avoiding the external splints.² He emphasized that in more proximal nerve lesions contractures are difficult to prevent owing to prolonged period of splinting and early tendon transfer may avoid this complication. Burkhalter’s proposition for early tendon transfer for wrist drop in radial nerve palsy is well adopted,² but to the best of authors’ knowledge, there is paucity of similar studies on early tendon transfer in surgical rehabilitation of hands with median nerve injuries. Since the single most important functional deficit in a median nerve palsy hand is the loss of thumb opposition, we carried out this study to evaluate the functional outcome of early tendon transfer for reconstructing thumb opposition.

2. Methods

This study was done prospectively and included 10 cases of isolated median nerve palsy due to traumatic or iatrogenic injury that underwent early tendon transfer for restoration of thumb opposition. Written informed consent was obtained from all the patients authorizing radiographic examination and clinical documentation. Institutional review board granted the approval for the study.

Neuropraxia was ruled out in closed injuries by waiting for 6 weeks for possible recovery. Only those with possible axonotmesis or neurotmesis of the median nerve in isolation were included in the study. Those with stiffness in trapezio-metacarpal joint of the involved side or a power less than Medical Research Council (MRC) grade 4 of the muscle to be transferred were excluded from the study. A detailed clinical history was taken and a meticulous clinical examination was performed making note of the cause of the injury, time since injury, whether dominant hand was involved, patient’s job profile, the functional deficit and the requirements narrated by each patient. Clinical and electro-diagnostic evaluation was done in selected cases only to look for any recovery since injury. A pre-operative course of occupational therapy was offered to all the patients to avoid or overcome joint stiffness especially in patients presenting late. A pre-operative evaluation in terms of power and precision grip strength and range of opposition of thumb was done in both normal and affected hand.

Grip Strength was measured with an automated dynamometer, using JTech GripTrack and PinchTrack (JTech Medical, Salt Lake City, United States). Power Grip was measured for five standard positions, representing objects of increasing size. An average of three readings, expressed in kilograms was used. Precision Grip was also measured in kilograms and included tip pinch, key pinch and palmar pinch.

Range of opposition was classified as per the Sundararaj and Mani scale³ as detailed: excellent {opposition to ring or little finger with interphalangeal (IP) joints extended}; good (opposition to index or middle finger tip with IP joint extended); fair (thumb IP joint flexes during opposition); and poor (no opposition restored).

2.1. Surgical Technique

All the procedures were performed under anaesthesia and tourniquet control. Extensor indicis proprious (EIP) opponensplasty was performed in nine patients. In eight of these, the tendon was passed over the ulnar border superficial to the flexor carpi ulnaris (FCU) (Fig. 1, Fig. 2, Fig. 3, Fig. 4, Fig. 5, Fig. 6). In one patient, EIP tendon was passed through the interosseous space. Palmaris longus (PL) opponensplasty was used in one patient due to scarring over the EIP tendon. The transferred tendon was attached to the insertion of abductor pollicis brevis (APB) with the wrist in neutral position and thumb in full abduction and opposition.

Fig. 1.

Intra-operative photograph showing extensor indicis proprious identification at wrist and 2nd metacarpo-phalangeal joint. Identification of EIP at extensor hood is important as it lies medial to the tendon of extensor digitorum slip of index finger.

Fig. 2.

Intra-operative photograph showing identification of EIP tendon at dorso-ulnar incision at the distal forearm.

Fig. 3.

Intra-operative photograph showing that EIP musculo-tendinous unit is freed and delivered through the dorso-ulnar forearm incision.

Fig. 4.

Intra-operative photograph showing rerouted EIP musculo-tendinous unit through ulnar border of forearm (superficial to flexor carpi ulnaris) volarly through a small incision over pisiform.

Fig. 5.

Intra-operative photograph showing criss-cross insertion of donor unit into the tendon of abductor pollicis brevis at the base of proximal phalanx.

Fig. 6.

Clinical result of a patient showing good range of active thumb abduction.

Median nerve exploration and repair was performed in four cases and was supplemented by a tendon transfer for thumb opposition in the same or different sittings. In one patient, the nerve was explored and tagged at the time of first surgery due to poor soft tissue condition and the repair was performed later.

The plaster slab was removed at 4 weeks post-surgery and active thumb abduction and opposition training was begun in consultation with the occupational therapist. Power grip, precision grip and opposition of thumb was measured on the injured and normal side at 4 weeks and thereafter at monthly intervals. Eight cases were followed up for 6 month and two cases for 12 months. Whether the patient was able to return to his original job or not and at what time was also noted.

2.2. Statistical Methods

The percentage deficit of strength between the normal and affected hand was calculated using the following formula:

$$\frac{(\text{Strength}\text{in}\text{normal}\text{hand})-(\text{Strength}\text{in}\text{affected}\text{hand})}{(\text{Strength}\text{in}\text{normal}\text{hand})}\text{×}100$$

The percentage increase in the grip strength was calculated using the following formula:

$$\frac{(\text{Post}-\text{operative}\text{strength}\text{at}6\text{months})-(\text{Pre}-\text{operative}\text{strength})}{(\text{Post}-\text{operative}\text{strength}\text{at}6\text{months})}\text{×}100$$

The preoperative grip strength of the affected and normal hand was compared using Mann-Whitney U test. The post-operative values were compared with the pre-operative values using the Wilcoxon signed rank test considering the non-parametric nature of the data. P < 0.05 was considered statistically significant.

3. Results

The median age at injury was 29 years (range; 8 years–57 years). Nine patients in the study were males. Seven patients had injury in the dominant hand. None of the patient had a bilateral injury. Six of the patients had open injury initially. Nine out of the ten patients had an associated fracture, 4 of which were open, with the paralysis of the median nerve. Another patient had a clean lacerated wound over the distal forearm with a median nerve transection. This includes a case of iatrogenic injury to median nerve following intercondylar humerus fracture fixation. Four patients had obvious neurotmesis and were repaired at the earliest. In one of these, the nerve could not be repaired in the first stage due to severely compromised tissue cover and had to be tagged and left alone. Nerve repair was planned at a later stage after the wounds were completely healed. Five out of six patients with open injuries had associated fractures. Two cases with open injuries, treated elsewhere initially, presented to us after wound healing with a positive Tinel’s sign and recovery of long finger flexors commensurate with the time since injury. In these cases spontaneous recovery was anticipated as they were considered to be axonotmesis lesions. An early tendon transfer alone was done pending nerve recovery. In due course of the study, they showed further signs of nerve recovery obviating the need for nerve exploration.

Median time for tourniquet application for EIP opponensplasty was 77 min (range; 63 min–122 min), and time taken for the only case of PL opponensplasty was 68 min. Nine patients underwent EIP transfer and in remaining patient PL opponensplasty was done.

3.1. Pre-operative comparison of affected hand with contralateral side

Maximum grip strength was found in position 2 in the affected as well as the normal sides and it decreased as the size of the object increased. The mean pre-operative power grip strength (all five positions) of the affected hand was only 23.83% as compared to the normal hand and this difference was statistically significant (p < 0.001). (Table 1).

Table 1.

Pre-operative power grip strength comparison of affected hand with contralateral side in 5 different positions (kgs = kilograms; pos = position; avg = average).

The mean pre-operative precision grip was tested in all the patients but 3 of the patients could not perform tip to tip, key, and palmar pinch due to ape thumb deformity. The maximum loss was observed in tip to tip pinch (14.68% strength as compared to normal side) as against 19.53% for key pinch and 18.27% for palmar pinch strength. The average precision grip strength was 17.35% as compared to the normal hand and this difference was found statistically significant (p < 0.001). (Table 2).

Table 2.

Pre-operative precision grip strength comparison of affected hand with contralateral side for tip to tip, key and palmar pinch.

3.2. Post-operative improvement in comparison to pre-operative status

Post operatively, we observed an increase in the grip strength as soon as at 4 weeks when the plaster was removed and the strength consistently improved with time on monthly assessment. The improvement in the power grip (all five positions) (Table 3, Table 4) and precision grip (tip, key and palmar) (Table 5, Table 6) of the operated hand was found statistically significant at 6 months follow-up (p < 0.001 for all except for palmar pinch where it was 0.05).

Table 3.

Graphic representation of the improvement in the power grip strength on monthly basis after surgery for all the 5 positions (kgs = kilograms; pos = position).

Table 4.

Improvement in the power grip strength on monthly basis after surgery for all the 5 positions depicting absolute values (pos = position; pre op = pre-operative; post op = post-operative).

	pos1	pos2	pos3	pos4	pos5
pre op	6.1	7.7	7.3	7.6	6.9
post op1	6.6	8	7.4	7.8	6.7
post op2	9.1	10.8	10.3	10.6	8.9
post op3	10.6	12.1	11.6	11.6	10.1
post op4	11.9	13.3	12.6	12.8	11.3
post op5	13.4	14.5	13.3	13.7	11.8
post op6	16.5	17.5	15.5	15.8	14.7
Increase (%)	170%	127%	112%	107%	113%

Table 5.

Graphic representation of the improvement in the precision grip (tip, key and palmar pinch) on monthly basis after surgery (kgs = kilograms).

Table 6.

Improvement in the precision grip (tip, key and palmar pinch) on monthly basis after surgery depicting absolute values (pre op = pre-operative; post op = post-operative).

	Tip	Key	Palmar
Pre op	0.99	1.09	1.44
post op1	2.04	2	1.99
post op2	2.24	2.35	2.14
post op3	2.43	2.53	2.37
post op4	2.64	2.68	2.55
post op5	2.75	2.9	2.64
post op6	3.2	3.2	2.8
Increase (%)	223%	193%	95%

The patients had statistically significant improvement in the percentage deficit (for all positions) at 6 months follow-up in comparison with the pre-operative percentage deficit values (p < 0.05).

We did not observe any tendon pull out in our study, neither any of our patients had an extensor lag following EIP transfer. None of our patients had poor opposition as per Sundararaj and Mani scale.³ Six patients who underwent EIP transfer had excellent opposition while 3 had good opposition. The patient, in whom PL opponensplasty was done, had an excellent opposition. The median time off work (MTW) in our study was 2.8 months (range; 1.5 months–3.4 months) with two patients not returning to work. One of them was 57-years-old manual labourer who decided to stop working considering it unsafe despite having good recovery. Another patient who underwent epineural median nerve repair at the wrist and EIP transfer through interosseous membrane was a hawker. He had persistent sensory loss and scarring. Thus, he was not comfortable in resuming the same work till last follow-up.

4. Discussion

Median nerve injury results in a serious functional set back to the patient. Since these nerve injuries occur more commonly in men who are physiologically active, athletic and in the most productive age group; this may cause considerable morbidity, if not managed efficiently.4, 5, 6 Despite a lot of scientific work in this field, the role of tendon transfer in upper extremity peripheral nerve injuries has long been debated and such transfers are largely utilized in irreparable lesions or when expected functional recovery does not occur following neurorrhaphy.

Median is a mixed nerve responsible for sensory innervation in the radial three and a half digits which contribute to the ever so important pinch area of the hand. It also supplies the intrinsic musculature involving complex movements of the thumb, thereby the prognosis following repair remains unpredictable. Brown reported the results of surgery in 272 upper extremity nerve injuries and found that neurorrhaphy of median nerve (38 repairs) resulted in partial functional recovery in half of the patients, whereas, other half had no functional benefit.⁷ He also suggested that tendon transfers for median nerve palsy have shown fairly good results in irreparable lesions of the nerve as 27 out of total 31 patients showed significant improvement.

It has also been observed that functional recovery in abductor pollicis brevis and opponens pollicis is usually lacking after neurorrhaphy in high median nerve injuries, even though extrinsic function returns in majority. If anomalous innervation does not exist, the loss of function of APB prevents optimal usage of the hand and may require external splintage in order to prevent development of contractures. The externally applied splints frequently impair remaining hand function and limit overall use of the limb preventing early rehabilitation.² The short opponens splint used in the management of the thumb following median nerve injury does not control rotational deformities of the thumb. With the loss of all pronation power, the supinator forces (extensor pollicis longus, adductor pollicis and deep head of the short flexors) still continue to act and may create a fixed supination deformity of the first metacarpal even with the use of the opponens splint.

An early tendon transfer seems to be more beneficial in high median nerve injuries due to the prolonged nerve regeneration time following repair. However it may also be useful in the low median nerve repairs as recovery in the intrinsic muscles of the thumb may be suboptimal in spite of a shorter distance to be travelled by regenerating axons to reinnervate the thenar muscles. Recovery following nerve repair is not only determined by the level of the injury but by a complex interplay of other factors like the time lag between injury and repair, crushing and contamination of the initial wound and age of the patient.⁸ Majority of our patients presented weeks or months after injury, some with secondarily healed scars suggesting a crushing mechanism with possible wound contamination and infection. Hence, we feel that it is wise to consider an early tendon transfer for both high and low median nerve palsy in such situations. Burkhalter, one of the foremost proponents of early tendon transfer, felt that such a transfer works not only as a substitute during the regrowth of the nerve, thereby improving function early, it also works as an internal splintage, providing early and a more accepted means of rehabilitation. Early tendon transfer following nerve repair or in the lesions in continuity (axonotmesis) provides early functional restoration and has the ability to prevent development of contractures and at the same time avoiding use of cumbersome external splints for long periods.² It adds power to the re-innervated muscles and in case the results of neurorrhaphy are poor, or in cases where the lesion is irreparable, it acts as a definitive treatment.

Loss of tip to tip pinch in median nerve palsy has been considered to be one of the most important contributor to problems in Basic Activities of Daily Living (BADL).⁹ Burkhalter emphasized that in order to utilize early tendon transfer appropriately, the transfer should focus on restoration of the single most important function lost, the transfer should not compromise with the remaining function of the hand, the transfer should not create a deformity if significant return of muscle function occurs following neurorrhaphy and the transfer should be a phasic one or one capable of phase conversion.²

It must be understood that, unlike radial nerve palsy where a tendon transfer procedure can be an alternate to nerve repair surgery, nerve repair is mandatory for restoration of hand sensations in median nerve injuries.

It has been suggested that thumb abduction and opposition, at times, may be preserved even after complete median nerve transection due to variability of thenar muscle innervation from an intact ulnar nerve. Also unilateral loss of opposition, especially in the nondominant hand, may result in a minimal functional loss not necessitating surgery.¹⁰ Since majority of our patients (7 out of 10) had their dominant hand involved and all our patients had a significantly compromised thumb opposition, they were enrolled for an early tendon transfer.

In our study, the choice of EIP as donor of choice was based on the success of the same in the available literature. Burkhalter et al. have reported excellent opposition after EIP opponensplasty in 57 out of 65 procedures that he performed.¹¹ Anderson et al. reported 87.5% excellent or good results following EIP opponensplasty in 40 hands.¹²

Cooney reported that flexor digitorum superficialis of the long and ring fingers are suitable for low median nerve palsy when strength was required.¹³ He preferred EIP transfer when only thumb mobility was desired. A comparison study of EIP and flexor digitorum superficialis (FDS) opponensplasty in 166 patients with median nerve paralysis of varied aetiology was done by Anderson et al. ¹² FDS was used for the transfer in 116 patients, whereas, EIP was used in rest 50. They showed that the EIP opponensplasty was best in supple hands and FDS opponensplasty was more suitable for less pliable hands. Baluch & Borschel performed the Camitz procedure at the time of median nerve decompression and/or reconstruction in 4 patients.¹⁴ All patients had excellent early return of function. But, Cooney proposed that the Camitz transfer is an abduction transfer rather than an opposition transfer and should be reserved for selected cases of long-term carpal tunnel syndrome.¹³ An ulnar nerve injury associated with a high median nerve injury limits the choice of available motors for opponensplasty.¹⁰ The flexor digitorum superficialis, palmaris longus and abductor digiti minimi are all not available and EIP is the obvious tendon of choice for an opponensplasty.

In our study, the pulp to pulp pinch was observed to have been restored immediately after slab removal in all cases and the pinch strength started improving quite earlier than the power grip strength. Schreuders et al. concluded that it is the pinch strength that significantly correlates with the strength of abduction and opposition of thumb.¹⁵ This means that although the transfer lacks the ability to provide a near normal power grip strength comparable to the normal hand, it could provide an excellent range of opposition, precision grip and a functionally useful hand.

The purpose of an early tendon transfer for thumb opposition is also served by the fact that MTW in our study was only 2.8 months. This was possible, in the present study, as the preceding nerve repair surgery facilitated sensory recovery. In a study on predictors for return to work in 96 patients who had undergone nerve repair surgery for median, ulnar, or combined nerve injuries, Bruyns et al. found that MTW was 7.3 months and only 59% patients returned to work within a year after injury.¹⁶ They found that grip strength loss, tip pinch strength loss, and sensory recovery differed strongly between the Return To Work (RTW) and no-RTW population. MTW was considerably less in our study, thereby, indicating that our approach puts the patient back to work in a very short period of time reducing the time off work.

Thus, it may be concluded that early tendon transfer has a unique role in the management of median nerve palsy hand. It acts as a substitute pending reinnervation of APB and acts as a helper to the recovered APB after its reinnervation following median nerve repair in neurotmesis or spontaneously in axonotmesis. However, the study suffers shortcomings in the form of small number of patients and limited period of follow-up.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Declaration of competing interest

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.