J Neurosurg (6 Suppl Pediatrics) 105:452–456, 2006

Collagen nerve guides for surgical repair of brachial plexus

birth injury

WILLIAM W. ASHLEY JR., M.D., PH.D., M.B.A., TRISHA WEATHERLY, P.A.-C.,

AND TAE SUNG PARK, M.D.

Department of Neurosurgery, St. Louis Children’s Hospital, Washington University School of Medicine,
St. Louis, Missouri

Object. Standard brachial plexus repair techniques often involve autologous nerve graft placement and neurotization.
However, when performed to treat severe injuries, this procedure can sometimes yield poor results. Moreover, harvesting
the autologous graft is time-consuming and exposes the patient to additional surgical risks. To improve surgical outcomes
and reduce surgical risks associated with autologous nerve graft retrieval and placement, the authors use collagen matrix
tubes (Neurogen) instead of autologous nerve graft material.
Methods. Between 1991 and 2005, the authors surgically treated 65 infants who had suffered brachial plexus injury at
birth. During this time, seven patients were treated using collagen matrix tubes (Neurogen). This study is a retrospective
analysis of the initial five patients who were treated using the tubes. Two patients underwent tube placement recently and
were excluded from the analysis because of the inadequate follow-up period.
Four of the five patients experienced a good recovery (motor scale composite [MSC] . 0.6), and three exhibited an
excellent recovery (MSC . 0.75) at 2 years postoperatively. The MSC improved by an average of 69 and 78% at 1 and 2
years, respectively. The movement scores improved to greater than or equal to 50% range of motion in most patients, and
the contractures were usually mild or moderate. Follow-up physical and occupational therapy evaluations confirm these
patients’ functional status. When last seen, four of five of these children could feed and dress themselves.
Conclusions. Technically, the use of the collagen matrix tubes was straightforward and efficient, and there were no com-
plications. The outcomes in this small series are encouraging.

KEY WORDS • brachial plexus • nerve graft • collagen nerve guide • peripheral nerve •
pediatric neurosurgery

HE incidence of birth-related brachial plexus palsy is
T approximately 0.5 to 2.5 per 1000 live births.3,5,7,10 Of
these infants, 70 to 95% will recover near-normal
function without intervention.7,10 However, a subset (10–
15%) may benefit from surgical exploration and repair.7 In
many cases, direct brachial plexus repair offers the best op-
portunity for functional motor recovery with minimal ortho-
pedic deformity. Standard brachial plexus repair techniques
often involve autologous nerve graft procedures, that is, su-
ral nerve graft retrieval and neurotization.9 However, autolo-
gous nerve graft placement and neurotization, when per-
formed in cases of severe injury, can sometimes yield poor
motor recovery. A particular problem of brachial plexus re-
pair in infants is that the sural nerves are not long enough to
provide an adequate graft when repair of the entire brachial
plexus is required. Moreover, harvesting the autologous
graft is time-consuming and exposes the patient to addition-
al surgical risks, including peroneal nerve injury, pain, loss
of function in nerve territory, and infection.4,6
Abbreviations used in this paper: MR = magnetic resonance;
MSC = motor score composite.
As our understanding of cellular and molecular neuro-
physiology increases, we have come to realize that physi-
cal continuity is only one of many critical factors involved
in successful nerve outgrowth and regeneration.4,6 Neuronal
growth must also be supported by neurotropic factors from
the distal stump. It has been shown that collagen and other
extracellular matrix proteins can act as a scaffold, enabling
directed neurite growth. Synthetic tubes constructed from
materials such as polyglycolic acid, polylactide-co-capro-
lactone, and silicone have been used as alternatives to au-
tologous graft placement, but in general, they have yielded
poor results.4,6 Tubes made of biological materials such as
laminin and collagen have been used with more success.1,6
Indeed, Neurogen collagen matrix tubes (Integra Neurosci-
ences, Plainsboro, NJ) have recently been approved by the
Food and Drug Administration for use as nerve graft mate-
rial based on studies in primates,1 but to our knowledge,
there have been no reports of their use in human brachial
plexus repair.
In an effort to improve outcomes and reduce the surgical
risks associated with autologous nerve graft placement, we
sought an alternative that would take advantage of our cur-

452 J. Neurosurg: Pediatrics / Volume 105 / December, 2006

Brought to you by Rutgers University | Unauthenticated | Downloaded 05/05/20 05:11 AM UTC
Collagen nerve guides for repair of brachial plexus birth injury

rent understanding of nerve regeneration and that would

significantly reduce the use of autologous nerves as graft
material. In this paper, we report the outcome of brachial
plexus repair in five infants in whom a collagen matrix tube
(Neurogen) was used as the graft material.

Clinical Material and Methods

Patient Population
Between 1991 and 2005, we performed brachial plexus
repair in 65 infants with birth-related brachial plexus injury.
In the latter part of this period (late 2002–2005), seven pa-
tients were treated with the placement of collagen matrix
tubes instead of autologous nerve graft material. This study
is a retrospective analysis of the initial five patients. The
last two patients were treated in January 2005 and were
therefore excluded from the analysis because of the inade-
quate follow-up period. The patients were chosen based on
the severity of the injury and the willingness of their fami-
lies to allow the placement of collagen tubes. We consid-
ered using collagen tubes in patients who had clinical evi-
dence of a very severe injury that we believed had a low
probability of recovering if traditional graft materials were
used. Before surgery, we discussed with the parents the pos-
sibility of using Neurogen tubes instead of autologous graft
material. If they consented to the use of Neurogen tubes,
then at the time of surgery, we made the final decision to
use the tubes based on the intraoperative findings (dis-
cussed later). We received institutional review board per-
mission for this study.
Table 1 shows the clinical data for the five patients in-
cluded in the study. The gestational age was approximately
39 weeks, and the infants’ average weight was approxi-
FIG. 1. Coronal T2-weighted MR image showing a pseudome-
mately 10 lbs. Most children had right-sided injuries with ningocele at the right C7–T1 level.
evidence of a pseudomeningocele visible on MR images
(Fig. 1). The average age at surgery was approximately 8
months. removed to expose the spinal nerve roots. Special care is
taken to avoid injury to the phrenic nerve.
Surgical Procedure In all five patients, the brachial plexus injury was severe
Our general techniques for brachial plexus exploration (Table 2). Extensive neuromas involving the spinal roots
and repair in infants have been previously described.8,9 and trunk portion of the brachial plexus were noted in all pa-
Briefly, after the induction of general anesthesia, needle tients, and in one most of the spinal roots were avulsed. Di-
electrodes are placed for the monitoring of somatosensory rect electrical stimulation proximal and distal to the neuro-
evoked potentials. The brachial plexus is explored through ma elicited minimal or no muscle contraction in the deltoid,
a supraclavicular approach, and the C5–T1 spinal roots, biceps, or triceps muscles or in the wrist and finger exten-
trunk, division of the brachial plexus, and axillary and sors and flexors. During brachial plexus repair in infants, we
phrenic nerves are exposed. The anterior scalene muscle is do not examine nerve action potentials across the neuroma
in the supraclavicular region because doing so would re-
quire extensive dissection involving sectioning of the pec-
TABLE 1 toralis muscles.
Our decision to use a collagen matrix graft is based on the
Clinical data for five patients with brachial plexus injuries*
severity of the injury, the size of the gap after removal of the
Case Gestational Birth Birth Age at Pseudo neuroma, and the diameters of the proximal and distal
No. Age (wks) History Weight (lbs) Op (mos) Side on MRI stumps. In addition to using collagen matrix tubes for short-
segment interposition grafting, we have also used them to
1 40 V 9.7 17.28 lt C-7
2 37 V 10.4 10.10 rt C7–T1
graft a single proximal stump (for example, C-5) to its dis-
3 41 V w/ S 12.4 5.97 rt no tal stump (C-5) and another nearby distal stump (C-6) with-
4 38 V w/ S 9.1 2.79 rt C7–8 in the same tube (also called a jump graft).
5 40 V w/ S 9.4 5.93 rt C-8 Once the decision to use collagen matrix tubes has been
mean† 39 6 2 10 6 1 865 made, we carefully assess the length of the gap and the di-
* Pseudo = pseudomeningocele; S = suction; V = vaginal. ameter of the nerve segment being replaced. Neurogen
† Values are expressed as means 6 standard deviations. tubes are available in various diameters (2–7 mm) and a sin-

J. Neurosurg: Pediatrics / Volume 105 / December, 2006 453

Brought to you by Rutgers University | Unauthenticated | Downloaded 05/05/20 05:11 AM UTC
 W. W. Ashley Jr., T. Weatherly, and T. S. Park

TABLE 2 imal and distal nerve stumps are placed within the ends of
Surgical data in five patients who the tube. The tube is secured with 7-0 Prolene suture that is
underwent brachial plexus injury repair* sewn to the epineurium in an interrupted fashion (Fig. 2).
Two to four sutures are placed circumferentially to secure
Case FU Dura- the tube. Once secured, the tube is filled with saline to re-
No. Surgical Findings Procedure tion (mos)†
move any air bubbles. For a jump graft, the diameter of the
1 C5–6 neuroma, severe C-5 neurolysis, C-5 inter- 26.36 tube must be wide enough to accommodate the proximal
upper trunk injury position graft, C5–6 stump of the involved root and the distal stumps of the in-
jump graft volved nerve and to jump over the recipient nerve. The col-
2 near-complete rupture C-5 & C-6 upper trunk 20.23
upper & middle trunks, graft, C-7 middle trunk
lagen tube is then placed and secured in a similar fashion.
moderate lower trunk graft, lower trunk neu-
injury rolysis Postoperative Care
3 severe panplexus disrup- C-5 upper trunk graft, 17.48 Patients wore a soft cervical collar and a brace to keep
tion C-7 middle trunk graft
4 ruptured upper trunk, long thoracic upper trunk 29.64 their elbow in flexion for 3 weeks. After the braces were re-
avulsed C5–T1 roots neurotization, C-5 mid- moved, the patients began physical therapy.
dle trunk graft, C-6
lower graft Data Analysis
5 extensive C5–8 injury upper/middle trunk neuro- 22.3
lysis, C-5 upper trunk We retrospectively reviewed the patients’ medical re-
graft, C-6 middle trunk cords and examined clinical, surgical, and follow-up data.
graft In another recent report,2 we defined the MSC as a numeric
* FU = follow-up.
measure for reporting the brachial plexus palsy motor out-
† The mean follow-up duration (6 standard deviation) was 23.2 6 4 come data. Briefly, upper-extremity muscle strength is grad-
months. ed on a four-point scale in each of nine muscle groups that
span the brachial plexus. A score of 1 represents trace or
no contraction; 2, visible movement; 3, movement against
gle length (2 cm). We used 5- to 7-mm-diameter tubes for gravity; and 4, movement against resistance. At the initial
discontinuities 2 cm or smaller. For a single interposition visit, if the strength grade is 4, then that muscle group is not
graft, the diameter must be wide enough to accommodate included in the MSC analysis. The strength in the involved
the proximal and distal nerve stumps. After the graft is groups is summed and reported as a proportion of the to-
soaked in saline, it is trimmed to the appropriate size and tal possible strength for those involved groups. This MSC
then placed in situ. Approximately 1 mm of both the prox- value ranges from 0.25 to 1.0. The MSC is calculated and

FIG. 2. Case 1. An intraoperative photograph showing the collagen tube in situ being placed between the C5–6 nerve
roots and upper trunk and secured with 7-0 Prolene suture.

454 J. Neurosurg: Pediatrics / Volume 105 / December, 2006

Brought to you by Rutgers University | Unauthenticated | Downloaded 05/05/20 05:11 AM UTC
Collagen nerve guides for repair of brachial plexus birth injury

recorded at each visit. By definition, the individual muscle TABLE 3

group strength is tracked by the MSC. An MSC of greater Outcome data in five patients who
than or equal to 0.6 represents good functional recovery, and underwent brachial plexus injury repair*
that of greater than or equal to 0.75 represents an excellent
Function
functional recovery. Joint
Case Gross Fine Mvmt Contrac-
Results No. Feeding Dressing Motor Motor ($50%)† tures

The mean duration of the surgical procedure was 4.5
hours, and there was minimal blood loss. The average hos-
pital stay was 2.2 days, and no patient experienced compli-
cations. Four of the five patients (Cases 1–3 and 5) exhibit-
ed a good recovery (MSC . 0.6), and three of five had
attained an excellent functional recovery (MSC . 0.75) at
the 2-year follow-up visit (Fig. 3). Using collagen tubes in-
stead of autologous grafts in these five patients resulted in
a mean increase in MSC by 69 and 78% at 1 and 2 years,
respectively. Joint movement scores improved to a range
of motion of at least 50% in most patients (Table 3), and
the contractures were usually mild or moderate. Moreover,
physical and occupational examinations performed during
the follow-up period confirmed the functional status. In-
deed, at the most recent follow-up visit, four of the five chil-
dren could feed and dress themselves. The first patient we
treated using Neurogen tubes has made remarkable prog-
ress. At just more than 2 years postoperatively, her recovery
is almost complete and she has nearly normal use of her left
arm (Fig. 4).
Discussion
Although safe and relatively uncomplicated, brachial
plexus repair does present some risks to the patient. The use
of autologous nerve graft material provides a good pathway
for the outgrowth of damaged nerves, but the incidence of
morbidity associated with brachial plexus surgery increases.
Indeed, bleeding, infection, and scarring are all risks asso-
ciated with harvesting autologous nerves. Moreover, ob-
taining the graft adds time and thus a greater risk of infect-
ion and increased cost. The use of artificial graft materials
such as collagen tubes may eliminate some of these risks.
Collagen nerve guides have been used successfully in ani-
mal models,1 but to our knowledge, this report is the first in
which collagen tubes were used for brachial plexus repair in
humans.
The use of collagen nerve guides significantly simplifies
the brachial plexus repair. The tubes come in a variety of
FIG. 3. Graph showing MSC data obtained in patients in whom
collagen tubes had been placed. Note that patients in Cases 1 to 3
did quite well and had an MSC greater than 0.75 at 2 years postop-
eratively. In Case 5 the patient recovered moderately well. In Case
4 the patient had a very severe injury and made no recovery. I = ini-
tial visit; P = preoperative visit.
1 yes yes AA good 15/16 none
2 yes yes AA delayed 10/16 moderate
3 yes yes AA delayed 9/16 moderate
4 no (flaccid) no (flaccid) flaccid flaccid 0/16 severe
5 yes yes AA good 12/16 mild (1
group)‡
* AA = age appropriate; mvmt = movement.
† A total of 16 joints were assessed. The numerator represents the num-
ber of joints that had at least 50% of movement.
‡ There were contractures in only one of the muscle groups evaluated.
diameters and a length of 2 cm. For now, we use the tubes
for gaps smaller than 2 cm. For neurite outgrowth to occur
successfully, current data suggest that the gap must be
small.1,4 Larger gaps result in aberrant growth or a lack of
growth. The diameter is also important. In addition to pro-
viding a physical scaffold for neurite extension, the space
within the tubes acts as a biochemical scaffold by providing
a channel for nerve growth factors produced by the proxi-
mal stump. Thus, a “good seal” on the proximal and distal
stumps ensures that these growth factors remain within the
tube. It also minimizes the number of stitches needed to se-
cure the tube; thus, suture-induced injury to the proximal
and distal stumps is minimized. Finally, removal of air in
the lumen of the tubes is critical. As mentioned, the tube
provides a channel for nerve growth factors produced by
the proximal stump. For nerve growth to occur, the tube
must fill with serum. The axons can then grow through this
nerve growth factor–enriched gellike matrix. Francel, et
al.,4 and others have suggested that axons will not grow
through an empty tube; they even propose placing a piece
of nerve within the tube to act as a nidus of nerve growth
factor production.
The results from the current study are encouraging. Al-
though we are unable to document histological evidence of
growth through these tubes, our clinical results suggest that
significant growth has occurred. It is of note that the studies
by Francel and colleagues4 were conducted in a rat model.
They used distances of up to 15 mm, which were propor-
tionately larger than the 2-cm gaps in humans. In addition,
these authors did not state the age of the rats they used. An
older age of the rats may have significantly affected the out-
come. We know from numerous studies that the nervous
system is more plastic at a young age. Finally, physical ther-
apy is difficult to model in rats. We attribute the success of
our procedures to three things. First, we carefully select the
diameter of the graft to ensure a good seal and concentra-
tion of growth factors in the tube. Second, we include a 1-
mm cuff of both the proximal and distal stumps in the tube.
These ends are the sites of growth factor production within
the tube. Finally, removing air from the tubes ensures a
good gel column for neurite growth and diffusion of growth
factors.
Because the current study comprises only a few patients,
we cannot offer strict criteria for the use of Neurogen tubes.

J. Neurosurg: Pediatrics / Volume 105 / December, 2006 455

Brought to you by Rutgers University | Unauthenticated | Downloaded 05/05/20 05:11 AM UTC

FIG. 4. Photograph showing remarkable functional improve-
ment in one of the patients treated with collagen tubes. She had a se-
vere left upper trunk injury and is now able to use her left arm al-
most normally.
However, in Table 4, we offer several guidelines based on
our current practice. As mentioned earlier, young animals
have a more plastic nervous system; thus, we currently use
the tubes to treat only birth-related injuries in young pa-
tients. Until we had proven the viability of the Neurogen
tubes, we wanted to limit the risk that the tubes would not
work as well as sural nerve grafts, ultimately reducing the
patient’s chance for optimal recovery. With this in mind, we
chose to use them to treat only the most severe injuries. We
used the tubes for gaps smaller than 2 cm and diameters
less than 5 mm because these were the sizes offered by the
company. However, as discussed earlier, smaller gaps and a
good fit are critical to encouraging maximal neurite out-
growth. Finally, at this early stage, we thought that we could
improve the chances of reproducible success and facilitate a
straightforward evaluation of the tubes when using them for
simple graft strategies. However, as our experience with the
tubes increases and tube design improves, the guidelines are
likely to be altered and include traumatic injuries, less se-
vere injuries, gaps larger than 2 cm, and more complex graft
strategies.
The outcomes in this small series of patients are encour-
aging with a rate of good functional recovery of 80%.
These outcomes are on par with those in other larger series
of brachial plexus repairs. Our results are particularly en-
couraging in light of the fact that all of the patients in this
group had severe injuries. When compared with surgical
results for severe brachial plexus injuries from our larger
series, the current group of patients seems to have fared
better. Only one patient in this group had a poor outcome.
In that case, the initial injury was very severe and the child
was tetraplegic. Intraoperative examination revealed a mas-
sive panplexus neuroma, and stimulation was nearly ab-
sent. Based on these findings, we were not surprised that
this child did not recover function. As shown in previous
studies, patients are unlikely to recover in the absence of at
least some muscle function.
TABLE 4
Summary of current guidelines for using Neurogen tubes
Guidelines for Using Neurogen Tubes
birth-related injury
severe injury (unlikely to improve using traditional techniques)
gap #2 cm
stump diameter 2–5 mm
simple grafting strategy
456
Brought to you by Rutgers University | Unauthenticated | Downloaded 05/05/20 05:11 AM UTC
W. W. Ashley Jr., T. Weatherly, and T. S. Park
We recognize that our study is small and that the follow-
up period is short. We plan to continue using collagen grafts
in selected patients and collecting outcome data. A prospec-
tive study would also be helpful to fully define the benefits
of using collagen grafts. For now, we use collagen nerve
guides for relatively simple small-segment interposition
grafts. More research in the field of nerve outgrowth may
allow us to extend the use of these tubes. In particular, by
enhancing the tubes with exogenous growth factors or by
incorporating strips of growth factor–impregnated collagen,
it may make it possible for us to use these nerve guides to
span large distances.
Conclusions
The use of the collagen matrix tubes was technically
easy and there were no complications. The outcomes in this
small series are encouraging. Four of the five patients made
a good recovery and were functional by 2 years postop-
eratively. Much more work needs to be done, but based on
this small case series, we suggest that collagen matrix tubes
are a safe alternative to autologous nerves for select short-
segment brachial plexus repairs.
References
1. Archibald SJ, Shefner J, Krarup C, Madison RD: Monkey me-
dian nerve repaired by nerve graft or collagen nerve guide tube.
J Neurosci 15:4109–4123, 1995
2. Ashley WW Jr, Park TS, Leonard J, Smyth M, Noetzel M, Wea-
therly T: Long-term motor outcome analysis following surgical
brachial plexus repair using a motor score composite, in Pro-
gram Book of the 34th Annual Meeting of AANS/CNS Sec-
tion on Pediatric Neurological Surgery, Orlando, Florida,
December, 2005. Rolling Meadows, IL: AANS/CNS, 2005, p 39
(Abstract)
3. Birch R, Ahad N, Kono H, Smith S: Repair of obstetric brachial
plexus palsy. Results in 100 children. J Bone Joint Surg Br
87:1089–1095, 2005
4. Francel PC, Francel TJ, Mackinnon SE, Hertl C: Enhancing nerve
regeneration across a silicone tube conduit by using interposed
short-segment nerve grafts. J Neurosurg 87:887–892, 1997
5. Haerle M, Gilbert A: Management of complete obstetric bra-
chial plexus lesions. J Pediatr Orthop 24:194–200, 2004
6. Midha R, Munro CA, Dalton PD, Tator CH, Shoichet MS: Growth
factor enhancement of peripheral nerve regeneration through a
novel synthetic hydrogel tube. J Neurosurg 99:555–565, 2003
7. Noetzel MJ, Park TS, Robinson S, Kaufman B: Prospective
study of recovery following neonatal brachial plexus injury. J
Child Neurol 16:488–492, 2001
8. O’Brien DF, Park TS, Noetzel MJ, Weatherly T: Management
of birth brachial plexus palsy. Childs Nerv Syst 22:103–112,
2005
9. Park TS, Kaplan SS: Birth brachial plexus injury, in Winn HR
(ed): Youmans Neurological Surgery, ed 5. Philadelphia: WB
Saunders, 2003, pp 3488–3498
10. Sherburn EW, Kaplan SS, Kaufman BA, Noetzel MJ, Park TS:
Outcome of surgically treated birth-related brachial plexus in-
juries in twenty cases. Pediatr Neurosurg 27:19–27, 1997
Manuscript received March 24, 2006.
Accepted in final form September 1, 2006.
Address reprint requests to: Tae Sung Park, M.D., Department of
Neurosurgery, St. Louis Children’s Hospital, One Children’s Place,
Suite 4S20, St. Louis, Missouri 63110. email: park@nsurg.wustl.edu.
J. Neurosurg: Pediatrics / Volume 105 / December, 2006