Elevated Body Mass Index Negatively Impacts Recovery of Shoulder Abduction Strength in Triceps Motor Branch to Axillary Nerve Transfers

Abstract

Background:

The purpose of this work was to evaluate the clinical outcomes of triceps motor branch to axillary nerve transfers and to identify prognostic factors which may influence these outcomes.

Methods:

A retrospective cohort included all patients who underwent a triceps motor branch to axillary nerve transfer (2010-2019) with at least 12 months of follow-up. The primary outcome measure was shoulder abduction strength assessed with British Medical Research Council (MRC) grade.

Results:

Ten patients were included with a mean follow-up of 19.1 (SD 5.9) months. Compared with preoperative MRC shoulder abduction strength (0.2 SD 0.4), patients significantly improved postoperatively (2.8 SD 1.6; P = .005). Increased body mass index (BMI) was significantly associated with worse postoperative MRC (P = .014).

Conclusion:

Triceps motor branch to axillary nerve transfer is a beneficial procedure for restoring shoulder function in patients presenting with either isolated axillary nerve or brachial plexus pathology. Patients with elevated BMI may not have as robust strength recovery and should be counseled carefully regarding prognosis.

Introduction

Axillary nerve injury resulting in deltoid paralysis yields marked disability to shoulder function. ¹ For patients that do not recover spontaneously by 6 months, 2 common surgical treatment options are either sural nerve interpositional grafting or a nerve transfer from the medial triceps motor branch of the radial nerve (triceps motor branch) to the axillary nerve.^2-7 However, increasing length of nerve grafts often results in suboptimal clinical outcomes. ⁸ Triceps motor branch to axillary nerve transfer was first described by Leechavengvongs et al ⁵ and involves transfer of the branch of the radial nerve innervating the long head of the triceps to the anterior branch of the axillary nerve. Previous literature has shown that most patients achieve greater than British Medical Research Council (MRC) grade 3 deltoid strength and greater than 90° shoulder abduction with this treatment.^9-12 Despite this, there remains a scarcity of information regarding patient selection and factors influencing outcomes of these nerve transfers. Our institution recently published data that demonstrated an association between increased body mass index (BMI) and worse motor outcomes in upper extremity nerve transfers. ¹³ Notably, subgroup analysis in that work demonstrated that increased BMI was associated with worse outcomes for shoulder reanimation but not for more distal transfers in the upper extremity. Given the relatively small sample size, a subsequent follow-up study was designed to further investigate the triceps motor branch to axillary nerve transfers. Thus, the purpose of this work was to evaluate the clinical outcomes of triceps motor branch to axillary nerve transfers and to identify prognostic factors which may influence these outcomes.

Methods

This study was conducted according to the Strengthening the Reporting of Observational Studies in Epidemiology guidelines.^14,15 Research ethics board approval was obtained to review the senior author’s electronic medical records. A retrospective cohort study was conducted for patients who underwent nerve transfer surgery from January 2010 to December 2019. Patients were included in the study if they received a triceps motor branch to axillary nerve transfer and had at least 12 months of follow-up. Patients who had concomitant nerve injuries or other nerve transfers were not excluded. Patients were excluded if they did not have extractable data in their medical records. Data were collected on patient demographics, comorbidities, nerve injury etiology, surgical management, and clinical evaluation. The primary outcome was shoulder abduction strength measured with MRC.^16,17 Two reviewers independently collected data and consensus was used to resolve disagreement. All clinical assessments were jointly performed in 1 peripheral nerve clinic by a plastic surgeon and a physiatrist. Electrodiagnostic studies were performed by a physiatrist and nerve transfers were performed by a single fellowship-trained peripheral nerve surgeon.

Statistical Analysis

Means and frequencies were used to report continuous and categorical data, respectively. Wilcoxon signed-rank test was used to compare changes in MRC scores. Spearman’s rho was used to calculate the correlation between BMI and postoperative MRC score. Mann-Whitney U test was used to determine the relationship between comorbidities and postoperative MRC score. A value of P < .05 was considered to be statistically significant.

Surgical Procedure

Our procedure is performed similarly to its first description by Leechavengvongs et al, ⁵ however modifying the procedure to use the medial triceps motor branch as our primary donor as described by Mackinnon. ¹⁸ With the patient in the prone position, a hockey stick incision is made on the posterior upper arm down to muscular fascia. The inferior border of the deltoid muscle is used to initiate dissection down to the quadrilateral space where the axillary nerve branches are identified. These recipient branches are then followed proximal to the motor branch to teres minor. The radial nerve branches to triceps are identified in the interval between the long and lateral heads of the triceps; branches to the long, medial, and lateral head of the triceps muscle are delineated using nerve stimulation. Under the operating microscope, the branch to the medial head of triceps is coapted to the anterior division of the axillary nerve proximal to the branch to teres minor. End-to-side nerve transfers are performed if the clinical context suggested that there may be a possibility of reinnervation via native axonal regrowth (eg, neuralgic amyotrophy, nerve in continuity), whereas end-to-end nerve transfers are performed when there was not a possibility of reinnervation via native axonal regrowth. Next, the posterior cutaneous branch of the axillary nerve is identified and neurolysed from the axillary nerve. Under the operating microscope, the posterior cutaneous branch of the axillary nerve is transferred in an end-to-side fashion to sensory component of the radial nerve to restore sensation over the deltoid.

Results

Ten patients undergoing 10 motor and nine sensory nerve transfers were included in the study (Table 1). Nine patients were male, and the mean age was 45.8 years (Table 2). The mean BMI was 28.9 kg/m²: four patients had a normal BMI (18.5-24.9), two patients were overweight (BMI 25-29.9), and four patients were obese (BMI >30). Smoking and hypertension were the two most frequent comorbidities, each being present in two patients.

Table 1.

Individual Patient Data.

No.	Age/sex	Injury type	Nerve injured	Comorbidities	BMI	ASA	Time from injury to surgery, mo	Coaptation type (motor transfer)	Coaptation type (sensory transfer)	Other nerve/tendon transfer	Preoperative MRC (shoulder abduction) grade	Postoperative MRC (shoulder abduction) grade	Follow-up time from surgery, mo
1	50/M	Low energy, closed injury	Brachial plexus	None	35.7	2	6.4	End-to-end	End-to-side	None	0	1	31.8
2	55/M	Low energy, closed injury	Axillary	None	39.0	3	6.3	End-to-end	End-to-side	None	0	1	18.1
3	34/M	Brachial neuritis	Brachial plexus	Smoker	22.2	3	7.9	End-to-end	End-to-side	None	0	4	19.5
4	69/F	Low energy, closed injury	Brachial plexus	None	19.7	2	7.3	End-to-end	End-to-side	None	0	4	15.4
5	18/M	Low energy, closed injury	Axillary	None	24.7	1	5.2	End-to-end	End-to-side	None	0	5	26.2
6	63/M	Low energy, closed injury	Brachial plexus	Smoker, HTN, CAD	37.3	3	6.2	End-to-end	End-to-side	None	0	1	20.9
7	49/M	High energy, closed injury	Axillary	None	26.5	2	3.8	End-to-end	End-to-side	None	1	4	12.6
8	59/M	High energy, closed injury	Brachial plexus	HTN	24.1	2	3.6	End-to-side	N/A	DFT (FCR to biceps, FCU to brachialis), ECRB to AIN	1	4	15.4
9	25/M	High energy, closed injury	Brachial plexus	None	25.8	2	3.4	End-to-end	End-to-side	None	0	1	17.5
10	36/M	High energy, closed injury	Axillary	None	33.9	2	3.7	End-to-side	End-to-side	None	0	3	14.0

Note. BMI = body mass index; ASA = American Society of Anesthesiologists; MRC = Medical Research Council; HTN = hypertension; CAD = coronary artery disease; DFT = double fascicular transfer; FCR = flexor carpi radialis; FCU = flexor carpi ulnaris; ECRB = extensor carpi radialis brevis; AIN = anterior interosseous nerve.

Table 2.

Patient Demographics (n = 10).

Sex a
Male	9 (90)
Female	1 (10)
Age, y b	45.8 (16.9)
BMI, kg/m2 b	28.9 (6.9)
ASA classification b	2.2 (0.6)
Comorbidities a
Active smoker	2 (20)
Hypertension	2 (20)
Coronary artery disease	1 (10)
Type II diabetes mellitus	0 (0)
Peripheral vascular disease	0 (0)

Note. BMI = body mass index; ASA = American Society of Anesthesiologists.

^an (percent).

^bMean (standard deviation).

Six injuries involved the brachial plexus, and four involved the axillary nerve in isolation. The most common injury etiology was trauma, seen in nine patients, followed by brachial neuritis, seen in one patient (Table 3). There were 10 triceps motor branch to axillary nerve transfers, of which eight were end-to-end and two were end-to-side (Table 4). There were nine concomitant axillary to radial sensory nerve transfers, all of which were end-to-side. One patient had an additional double fascicular transfer and extensor carpi radialis brevis to anterior interosseous nerve transfer. Patients underwent surgery at a mean of 5.4 months following injury or onset of symptoms. The mean follow-up time was 19.1 months.

Table 3.

Peripheral Nerve Injury Clinical Data (n = 10).

Nerve injury a
Brachial plexus	6 (60)
Axillary	4 (40)
Etiology a
Trauma	9 (90)
Low energy, closed injury	5 (50)
High energy, closed injury	4 (40)
Brachial neuritis	1 (10)

^an (percent).

Table 4.

Nerve Transfer Clinical Data (n = 19).

Nerve transfer a
Triceps to axillary (motor)	10 (52.6)
End-to-end	8 (42)
End-to-side	2 (10.5)
Axillary to radial (sensory)	9 (47.4)
End-to-side	9 (47.4)
Time from injury to surgery, mo b	5.4 (1.7)
Follow-up time from surgery, mo b	19.1 (5.9)

^an (percent).

^bMean (standard deviation).

Compared with preoperative MRC shoulder abduction strength (0.2 SD 0.4), patients significantly improved postoperatively (2.8 SD 1.6; P = .005) (Table 5). There was no significant difference in mean postoperative MRC grade between end-to-end (2.6 SD 1.8) and end-to-side nerve transfers (3.5 SD 0.7, P = .71). When stratified into normal BMI (18.5-24.9), overweight (BMI: 25-29.9), and obese (BMI: >30), patients had mean postoperative MRC grades of 4.3, 2.5, and 1.0, respectively. Increased BMI was significantly associated with lower postoperative strength (P = .014), but age and time from injury to surgery were not found to be statistically significant (Table 6). The presence or absence of comorbidities did not have any effect on postoperative strength (Table 7).

Table 5.

Nerve Transfer Outcomes (n = 10).

Movement for assessment	Preoperative MRC grade a	Postoperative MRC grade a	P value b	Change in MRC grade a
Shoulder abduction (deltoid)	0.2 (0.4)	2.8 (1.6)	.005*	2.6 (1.5)

Note. MRC = Medical Research Council.

^aMean (standard deviation).

^bWilcoxon signed-rank test.

^*Statistical significance at P < .05.

Table 6.

Correlation Between Patient Demographics and Postoperative MRC grade Following Nerve Transfer (n = 10).

Patient characteristic	Correlation coefficient a	P value a
BMI	−0.743	.014*
Age	−0.226	.530
Time from injury to surgery	0.065	.859

Note. MRC = Medical Research Council; BMI = body mass index.

^aSpearman rho.

^*Statistical significance at P < .05.

Table 7.

Relationship of Comorbidities on Postoperative MRC grade Following Nerve Transfer (n = 10).

Comorbidity	Comorbidity present		Comorbidity absent
	n	Postoperative MRC grade a	n	Postoperative MRC grade a	P value b
Active smoker	2	2.5 (2.1)	8	2.875 (1.6)	1
Hypertension	2	2.5 (2.1)	8	2.875 (1.6)	1
Coronary artery disease	1	1 (–)	9	3 (1.6)	.5

Note. MRC = Medical Research Council.

^aMean (standard deviation)

^bMann-Whitney U test.

Discussion

The purpose of this work was to evaluate the clinical outcomes of triceps motor branch to axillary nerve transfers and to identify prognostic factors which may influence these outcomes. Among a cohort of 10 patients with a mean follow-up of 19.1 months, patients’ shoulder abduction strength significantly improved following triceps motor branch to axillary nerve transfers. Increased BMI was significantly associated with poorer postoperative strength, but age and time from injury to surgery were not found to be correlated with postoperative strength.

There are several studies that are concordant with our findings that triceps motor branch to axillary nerve transfer can successfully provide reanimation of shoulder abduction.^12,19 Desai et al ²⁰ reviewed a cohort of 27 patients who underwent triceps motor branch to axillary nerve transfers and found that 22 achieved at least MRC grade 3 strength postoperatively, and the average shoulder abduction was 114° compared with 12° preoperatively (P < .05). Wolfe et al ²¹ also described their triceps motor branch to axillary nerve transfers on 14 patients with axillary nerve palsy and found that 12 achieved at least MRC grade 3 strength, while shoulder abduction increased from 6° to 85° (P < .001). Bertelli et al ²² presented a series of three patients (mean age: 22.7 years) with isolated axillary nerve injury who all improved to MRC grade 4 deltoid strength at their 18-month follow-up following surgery, which was performed at a mean time of nine months after injury. The triceps motor branch to axillary nerve transfer has also been combined with other nerve transfers to yield beneficial results in upper brachial plexus injury patients.^6,7

Furthermore, there is preliminary clinical evidence of the relationship between BMI and outcomes following nerve transfers. In Lee et al’s retrospective study of 21 patients undergoing triceps motor branch to axillary nerve transfers, they identified BMI as a factor which negatively affected postoperative deltoid muscle strength. They speculated that high BMI patients with heavier arms may not also have larger deltoid muscles, and thus would not have the additional strength to move their arm. ¹² Conversely, Yang et al ¹⁹ examined nine patients who underwent triceps motor branch to axillary nerve transfers and found no relationship between BMI and shoulder abduction range of motion or Disabilities of the Arm, Shoulder, and Hand score.

With respect to other upper extremity nerve transfers, Sallam et al ²³ examined a cohort of 55 patients undergoing nerve transfers for radial, ulnar, and median nerve injuries, and found a downward trend in MRC grade with increased BMI. Socolovsky examined 18 patients who underwent spinal accessory to suprascapular nerve transfer in brachial plexus injuries and demonstrated that elevated BMI negatively impacted recovery of shoulder abduction. ²⁴ However, Socolovsky’s second study of 40 patients undergoing Oberlin’s procedure demonstrated no impact of BMI. ²⁵ In addition, Texakalidis et al ²⁶ showed no influence of BMI on combined nerve transfer outcomes for shoulder reanimation, and Xiao et al ²⁷ showed no influence of BMI on outcomes of intercostal nerve transfers for elbow flexion.

Other factors that have been shown to contribute to the clinical success of nerve transfer surgery include time to surgery and age. Lee et al ¹² found that deltoid strength negatively correlated with delay from injury to surgery (Spearman rho = −0.533, P = .013) and that average deltoid strength was significantly decreased if nerve transfer surgery was delayed for more than nine months (2.7 SD 0.8, P = .02). In a series of 121 patients with isolated and combined axillary nerve injuries receiving neurolysis, grafting, or nerve transfer, Bonnard et al ² demonstrated that a delay of at least 5.4 months in nerve grafting results in a lower final shoulder strength outcome (P < .003). Terzis and Barmpitsioti ²⁸ reported that patients with a time to surgery of less than four months had significantly improved shoulder function after axillary nerve reconstruction with either nerve transfer or nerve grafts compared with a time to surgery of greater than 8 months (64.60 SD 28.64 vs 47.40 SD 26.41 degrees of shoulder abduction, P < .02). Unfortunately, our study was not adequately powered to meaningfully examine this relationship.

Bonnard et al, ² Lee et al, ¹² and Wehbe et al ²⁹ have all collectively demonstrated that patient’s age is one of the contributing factors on the outcome of nerve surgery procedures, with increasing age resulting in worse clinical outcomes. In Lee et al’s ¹² previously mentioned study of 21 patients, they found that deltoid strength negatively correlated with patient age (Spearman rho = −0.585, P = .005). They reported an average deltoid MRC grade of 3.9 SD 0.7 in patients aged 39 years or younger, which was significantly higher than in patients between 40 and 49 years old (2.9 SD 1.0, P = .04). They did not observe any useful recovery of deltoid strength in patients older than 50 years. When compared with our study, we demonstrate that patients who achieved deltoid MRC grade 4 or higher ranged from 18 to 69 years of age. Wehbe et al ²⁹ performed a retrospective study of 33 patients with axillary nerve injuries who underwent neurolysis or nerve grafting, and found a higher proportion of favorable outcomes (MRC grade 3 or higher) in patients younger than 25 years (8/14) compared with patients aged 25 years or older (8/19), although this did not achieve statistical significance (P < .1). In Bonnard et al’s ² previously mentioned study of 121 patients, they found a significant downward trend in the proportion of deltoid MRC 4 to 5 results as patient age increased (P = .025). Our study did find a negative correlation between age and postoperative MRC; however, due to low sample size, we were not able to reach statistical significance.

Limitations

This study is limited primarily by the low sample size, which limited its power and restricted the possibility of exploring the impact of several comorbidities on axillary motor nerve transfer outcomes. The second notable limitation is with respect to the minimum of 12 months of follow-up for inclusion, which may not be sufficient to observe the final recovery of patients following nerve transfer surgery. In addition, due to previously mentioned published results which indicate poor outcomes in delayed surgical intervention, we chose to operate within eight months after injury. Consequently, our lack of patients with delayed surgical intervention limits our ability to effectively comment on time to surgery as a predictor of strength recovery.

Conclusions

This retrospective study adds to the growing body of literature demonstrating that triceps motor branch to axillary nerve transfer can be used for shoulder reanimation in axillary nerve injuries. Notably, patients with elevated BMI may not have as robust strength recovery following triceps to axillary motor nerve transfer—this should be an important consideration in patient selection and be appropriately integrated into patient counseling regarding prognosis.