RETRACTED

ORIGINAL COMMUNICATION

Multiple Lumbar Plexus Neurotizations of the Ninth,Tenth, and Eleventh Intercostal Nerves

R. VIALLE,1–3* C. COURT,4 IAN HARDING,5 J.F. LEPEINTRE,2 N. KHOURI,3 AND M. TADIE2

1Ecole de Chirurgie de l’Assistance Publique des Hopitaux de Paris, Paris, France2Department of Experimental Neurosurgery, Kremlin-Bicetre Hospital, Kremlin-Bicetre, France

3Department of Paediatric Orthopaedics, Armand Trousseau Hospital, Paris, France4Department of Orthopaedic Surgery, Kremlin-Bicetre Hospital, Kremlin-Bicetre, France

5Department of Orthopaedic Surgery, Nuffield Orthopaedic Centre, Oxford, United Kingdom

The topographic anatomy of the lower intercostal nerves is less known than the upper onesexcept for the twelfth intercostal nerve. It is possible to use the lower intercostal nerves to carryout a neurotization of the lumbar roots. We studied the anatomy of the ninth, tenth, and eleventhintercostal nerves in order to specify the data of descriptive and topographic anatomy allowingto carry out their harvesting under good conditions. Ninth, tenth, and eleventh intercostal nervesof 50 cadavers were dissected. The proximal part of the nerve in the posterior intercostal spacewas exposed through a posterior approach. The lateral intercostal space was exposed through alateral approach, under the latissimus dorsi, which made it possible to harvest the intercostalsnerve. The proximal course of the nerve in posterior intercostals space was the same in all thecases. The nerve moves obliquely towards the outside to reach the lower border of the rib. Theexit of posterior intercostal space is a fibrous strait that marks the entry of a channel betweentwo muscular layers. We describe an aponevrotic channel in which the nerve and the vesselsare, immediately at the lower border of the cranial rib. The mean total length of intercostal nerveharvested by our technique was 17.96 cm for the ninth intercostal nerve, 17.14 cm for the tenthintercostal nerve, and 15.94 cm for the eleventh intercostal nerve. The bifurcation of the inter-costal nerve in a deep branch and the ramus cutaneus lateralis was found in the majority of thecases, from 9.5 to 21 cm of the emergence of the intercostal nerve in posterior intercostal space.This anatomical study of the ninth, tenth, and eleventh intercostal nerves in posterior intercostalspace and lateral intercostal space appears to us to allow the realization of a reliable surgical har-vesting. Clin. Anat. 19:51–58, 2006. VVC 2005 Wiley-Liss, Inc.

Key words: lower intercostal nerves; lumbar roots neurotization; surgicalharvesting; anatomical study

INTRODUCTION

The 12 ventral rami of the thoracic nerves are

called intercostal nerves and there are sensory and

motor fibres within these intercostal nerves. Asfaza-

dourian et al. (1999) have described the neurotiza-

tions of injured brachial plexus roots using them

when it is often necessary to take three to four inter-

costal nerves and to transfer them to the axillary

fossa. The topographic anatomy of the upper inter-

costal nerves (from the first to the sixth) was previ-

ously well described by Davies et al. (1932) and

Williams et al. (1989). The ventral ramus of the sec-

ond, third, fourth, fifth, and sixth thoracic nerves, and

the small branch from the first thoracic, are confined

to the parietes of the thorax, and are named thoracic

intercostal nerves. They pass anteriorly in the inter-

costal spaces adjacent and distal to the intercostal

vessels. At the back of the chest they lie between

the pleura and the ligamentum intercostalis internus(also called posterior intercostal membrane), piercing

the latter, running between the two planes of inter-

costal muscles as far as the middle of the rib.

*Correspondence to: Dr. Raphael Vialle, 105 avenue Andre Mori-

zet, 92100 Boulogne Billancourt, France. E-mail: ravialle@noos.fr

Received 9 May 2004; Revised 21 December 2004; Accepted 3

February 2005

Published online 26 September 2005 in Wiley InterScience (www.

interscience.wiley.com). DOI 10.1002/ca.20148

VVC 2005 Wiley-Liss, Inc.

Clinical Anatomy 19:51–58 (2006)

This article has been retracted because the authors and the editor of Clinical Anatomy have agreed that the article is nearly identical to the article

published in Surgical and Radiologic Anatomy 27(1):8-14, 2005.

[Retracted Article]

RETRACTED

Walmsley (1916) described the lateral cutaneousbranches (rami cutanei laterales) that are derived fromthe intercostal nerves, approximately midwaybetween the vertebra and sternum. These lateralcutaneous branches pierce the musculus intercostalisexternus and serratus anterior and divide into anteriorand posterior branches. The anterior branches runforward toward the anterolateral aspect of the chest,supplying skin. In addition, the anterior branches ofthe fifth and sixth nerves supply the upper digita-tions of the obliquus externus abdominis. The posteriorbranches course posteriorly supplying the skin overthe scapula and latissimus dorsi.

The ventral ramus of the seventh, eighth, ninth,tenth, and eleventh thoracic nerves continue anteri-orly from the intercostal spaces into the abdominalwall. They are named thoracoabdominal intercostalnerves in the description by Williams et al. (1989).Anatomically, they exhibit the same characteristicsas the more proximal nerves until they reach theanterior aspect of the intercostal space, when theypass behind the costal cartilages, between the obli-quus internus and transversus abdominis, into thesheath of the rectus abdominis that they supply. Theythen terminate as anterior cutaneous abdominalbranches, supplying the skin in this region. Studiesby Sakamoto et al. (1996a,b) and Schalow et al.(1992) have shown that the lower intercostal nervessupply the intercostal and abdominal muscles, the

last three send branches to the serratus posterior infe-rior. At approximately the middle of their course

they give off lateral cutaneous branches that pierce

the musculus intercostalis externus and the obliquusexternus abdominis (in the same line as the ramus cuta-neus lateralis of the upper thoracic nerves) and then

divide into anterior and posterior branches (that sup-

ply the skin of the abdomen and back). The ramuscutaneus lateralis supplies the digitations of the obli-quus externus abdominis, and extends downward and

forward nearly as far as the lateral edge of the rectusabdominis, whereas the posterior branches pass back-

ward to supply the skin over the latissimus dorsi.It is possible to use the lower intercostal nerves to

carry out a neurotization of the lumbar roots as pro-

posed by Zhao et al. (1997). The experiments carried

out on animals by Tok et al. (1991) and on cadavers

by Hauge (1991) and Lim and Baskaran (2001)

describe only a limited number of cases. It is inaccu-

rate, as well as an oversimplification, to transpose the

known clinical anatomy of the higher intercostal

nerves to that of the lower intercostal nerves and the

aim of this study therefore, is to more precisely

define the anatomy of the ninth, tenth and eleventh

intercostal nerves. We have studied 50 cadavers to

describe and clarify the topographic anatomy that will

facilitate their harvesting if neurotisation is indicated.

MATERIALS AND METHODS

Thirty-seven embalmed and thirteen fresh speci-

mens (28 males, 22 females) of undetermined age

were dissected by the same author (R.V.). The

thoracic and lumbar spine was exposed through a

median posterior approach. The ninth, tenth, and

eleventh intercostal nerves were localized at their

emergence from the intervertebral foramen. The

intercostal arteries and veins were also dissected and

identified. Their position relative to the other neuro-

vascular structures was noted as well as their location

in the posterior intercostal space. The initial course

of the nerve, through the posterior intercostal space,

was visualized by this posterior approach. Right and

left latissimus dorsi muscles were exposed through

two lateral incisions, detached and taken medially to

visualize the ninth, tenth and the eleventh intercos-

tal spaces. The musculus intercostalis externus was

detached from the caudal edge of the higher rib to

expose the intercostal vein, artery and nerve in lat-

eral intercostals space. The ramus cutaneus lateralis ofthe intercostal nerve was isolated and sectioned just

after its exit from the musculus intercostalis externus.The dissection of the deep medial branch of the

intercostal nerve was continued as distally as possi-

Fig. 1. Anatomic dissection of a fresh female cadaver. Posterior

view of the eleventh posterior intercostal space. After careful opening

of the ligamentum intercostalis internus, intercostal nerve and vessels

are well-visualized, adjacent to the posterior parietal pleura.

52 Vialle et al.

RETRACTED

ble. For each nerve, the overall length of the har-

vested nerve and the level of the bifurcation

between the deep medial and the ramus cutaneus lat-eralis were measured. Measurements were taken

from the emergence of the nerve from the interver-

tebral foramen and the diameter of each nerve was

recorded. For each cadaver, the length of the trunk

was evaluated by the distance between the spinous

process of the seventh cervical vertebra and the pos-

terior superior iliac spine. The nerve length neces-

sary to perform an intradural lumbar root neurotiza-

tion was noted. The length was measured between

the nerve exit from the intervertebral foramen to the

nerve roots under the conus medullaris. We calculated

the difference between the length of harvested

nerve and the necessary length required at each

level to assess neurotization feasibility.

All parameters were collected in a computerized

database and analyzed by the use of the SPSS1 soft-

ware (SPSS Inc., Chicago, IL). The population dis-

tribution was assumed to be normal regarding the

trunk size. Statistical analysis was carried out using

Pearson correlation coefficient (R). Correlations

between non-binary and binary numeric variables

were made by the use of an unpaired t-test for

equality of means and a one-way ANOVA. Differen-

ces with P-values <0.05 were considered statistically

significant. Because of the distribution, statistical

analysis regarding the gender was carried out using

non-parametric tests.

RESULTS

The mean distance between the spinous process of

the seventh cervical vertebra and the posterior supe-

Fig. 3. Difference between

the harvested ninth intercostal

nerve lengths and the theoretical

necessary lengths to perform intra-

dural nerve root neurotization. The

harvested length was sufficient in

the 300 harvesting procedures.

Fig. 2. Anatomic dissection of a fresh male cadaver. Posterior

view of the tenth posterior and lateral intercostal space. After

removing the musculus intercostalis externus, intercostal nerve and

vessels are well-visualized, in an aponeurotic channel, just at the

lower border of the tenth rib. Note the nerve bifurcation into a deep

and superficial branch.

53Anatomical Study of the Lower Intercostal Nerves

RETRACTED

rior iliac spine was 48.56 cm (40–58 cm). This value was

significantly higher in men than in women (P < 0.001

Wilcoxon’s test). The location of nervous and vascular

elements in the posterior intercostal space was the

same in all the cases. The intercostal nerve and ves-

sels were always in contact with the parietal pleura, in

a space filled with fatty tissue and separated from the

intervertebral and vertebrocostal muscles by a thin

membrane, the ligamentum intercostalis internus (also

called posterior intercostal membrane).

The dorsal ramus of the thoracic nerve passed

through a defect in the medial and cranial part of

the ligamentum intercostalis internus. This dorsal ramusthen perforated the musculus erector spinae (that it

supplies), the musculus trapezius and the other super-

ficial muscles of the back to reach the fascia superfi-

Fig. 4. Difference between

the harvested tenth intercostal

nerve lengths and the theoretical

necessary lengths to perform

intradural nerve roots neurotiza-

tion. The harvested length was

sufficient in 299 of the 300 har-

vesting procedures.

Fig. 5. Difference between

the harvested eleventh intercostal

nerve lengths and the theoretical

necessary lengths to perform

intradural nerve roots neurotiza-

tion. The harvested length was

sufficient in 298 of the 300 har-

vesting procedures.

54 Vialle et al.

RETRACTED

cialis and the skin. The course of the nerve through

the posterior intercostal space was the same for the

ninth and the tenth intercostal nerves. In all our

specimens the course of the nerve commenced crani-

ally within the posterior intercostal space. The nerve

then coursed laterally to reach the lower (caudal)

border of the rib. Without exception, the vessels

were cranial to the emergence of the nerve, adjacent

to the lower border of the rib. The ninth, tenth, and

eleventh intercostals nerves took the same course in

the posterior intercostal space. The emergence of

the eleventh in the posterior intercostal space was

more distal, always in the caudal half of posterior

intercostal space (Fig. 1). The eleventh intercostal

nerve was uniformly further away from the vessels

of the posterior intercostal space than its counterpart

in the tenth and the ninth spaces.

Following the exit of the nerves from the poste-

rior intercostal space, they run forward along the

thoracic wall between the musculus intercostalis inter-nus and the musculus intercostalis externus. In the elev-

enth intercostal space, the intercostal nerve and the

vessels were separated each from the other by a

fibrous arcade lying between the deep and superficialintercostalis musculi (Fig. 2). This arcade was not

found in the ninth and the tenth intercostal space.

Through the lateral intercostal space, the course of

the intercostal nerve and vessels was consistently

close to the caudal edge of the cranial rib. In all

specimens, the intercostal nerve was found easily

after detachment of the musculus intercostalis externusfrom the lower border of the cranial rib. In all speci-

mens, the nerve and the vessels were in the same

space, between the deep and the superficial intercostalmusculi. This space was defined by a fine fascia,

forming a tunnel under the rib.

The mean total length of intercostal nerve har-

vested was 17.96 (10–27 cm) cm for the ninth inter-

costal nerve, 17.14 cm (10–20 cm) for the tenth

intercostal nerve and 15.94 cm (10–25 cm) for the

eleventh intercostal nerve. The length of harvested

nerve was not correlated with the distance between

the spinous process of the seventh cervical vertebra

and the posterior superior iliac spine.

The difference between the length of harvested

nerve and the nerve length necessary to perform

lumbar roots neurotization was positive in 297 of the

300 cases of nerve harvesting (Figs. 3–5). In two

cases of ninth intercostal nerve harvesting and in

one case of tenth intercostal nerve, the length was

insufficient to perform a lumbar root neurotization in

‘ideal’ conditions (Fig. 6). The data about harvested

TABLE 1. Main Results of the Intercostals Nerve Lengths*

nMinimum

(cm)Maximum

(cm)Mean(cm)

SD(cm)

Trunk size 50 40 58 48.56 4.13T9 harvested length 100 10 27 17.96 2.50T10 harvested length 100 10 20 17.14 2.06T11 harvested length 100 10 25 15.94 2.50Theoretical necessary length for T9 neurotization 100 6.5 13.5 11.25 1.54Theoretical necessary length for T10 neurotization 100 4.2 10.5 8.41 1.62Theoretical necessary length for T11 neurotization 100 1.8 7.6 5.15 1.45

*Harvested and theoretical, measured on 50 anatomical subjects.

Fig. 6. Anatomic dissection of a fresh male cadaver. Posterior

view after bilateral harvesting of the ninth, tenth and eleventh inter-

costals nerves. The length of nerve harvested by means of one sec-

ond incision at the lateral border of the latissimus dorsi muscle

seemss sufficient to us to carry out a nervous anastomosis on the

level of the conus medullaris.

55Anatomical Study of the Lower Intercostal Nerves

RETRACTED

intercostal nerve lengths and theoretical values are

summarized in Table 1.

The division of the intercostal nerve into two

branches was found in 89 of the 100 ninth intercostal

nerves, 80 of 100 tenth intercostal nerves and 82 of

100 eleventh intercostal nerves (Fig. 7). For the

ninth intercostal nerve, the bifurcation was located

on average at 16.19 cm (10–21 cm) from the emer-

gence of the nerve in the posterior intercostal space.

For the tenth intercostal nerve, the bifurcation was

located on average at 15.08 cm (11–19 cm) from the

emergence of the nerve in the posterior intercostal

space. For the eleventh intercostal nerve, the bifur-

cation was located on average at 13.20 cm (9.5–19 cm)

from the emergence of the nerve in the posterior

intercostal space. The distance between the emer-

gence of the nerve and the bifurcation was not stat-

istically higher in men than in women and it was not

statistically higher in cadavers with the longest

trunks. The distance between the emergence of the

eleventh intercostal nerve and its bifurcation was

significantly positively correlated with the distance

between the spinous process of the seventh cervical

vertebra and the posterior superior iliac spine (r ¼0.298 with P ¼ 0.006 Pearson’s test).

DISCUSSION

The topographic anatomy of the thoracic intercos-

tal nerves (ventral ramus of second, third, fourth, fifth

and sixth thoracic nerves) has been well described.

The topographic anatomy of the lower intercostal

nerves is less well known. Only the twelfth intercos-

tal nerve has been previously well described by

Aubert et al. (1981). Its large diameter and signifi-

cant contribution to the innervation of the abdomi-

nal wall is the base of sensitive cutaneous flaps

reported by Yan and Zhong (2000).

Thoracoabdominal nerves have the same anatomi-

cal arrangement as the upper ones in the lateral

intercostal space. The ninth, tenth and eleventh

nerves are further away from the lower border of the

cranial rib in the initial portion of their course than

the upper thoracic intercostal nerves. In contrast

with Hardy (1988), we consistently found the same

anatomical pattern of the intercostal nerves in the

posterior intercostal spaces. Their course is more

oblique, running in a proximal and lateral direction,

reaching the lower border of the cranial rib before

the exit from the posterior intercostal space. The

knowledge of this anatomy of the proximal portion

of the lower intercostal nerves in the posterior inter-

costal space makes it possible to approach it surgi-

cally without risking injury to it. The careful open-

ing of the ligamentum intercostalis internus (the poste-

rior intercostal membrane) makes it possible to

easily identify the nerve, which crosses obliquely in

a fatty space. It is necessary to be particularly careful

to avoid injury of the posterior parietal pleura,

located just under the nerve. The exit from the pos-

Fig. 7. Anatomic dissection

of a fresh female cadaver. Lateral

view of the ninth intercostal space

under the latissimus dorsi muscle.

The deep and superficial branches

of the ninth intercostal nerve are

well-visualized just under the cau-

dal border of the ninth rib (white

points).

56 Vialle et al.

RETRACTED

terior intercostal space is along a layer of fibrous tis-

sue that marks the entry of a channel between two

muscular layers. In contrast with the traditional ana-

tomic description by Rouviere (1967) of a wide free

area between the deep and superficial intercostal

muscle, we have described an aponeurotic channel,

just at the lower border of the cranial rib, in which

the nerve and the vessels lie. The nerve and the

vessels are located in the same space, the nerve

being always the most caudal element. A fibrous

septum separates the eleventh intercostal nerve from

the intercostal artery and vein leaving the posterior

intercostal space. Thereafter, the course of vascular

and nervous elements continues in the same anatom-

ical space.

The bifurcation of the intercostal nerve into a

deep branch and the ramus cutaneus lateralis was

found in the majority of the cases. Whenever we did

not find it, it is probable that the superficial branch

was replaced by numerous fine branches (rami cuta-nei laterales) of the nerve that could not be found

due to their small size. When the bifurcation was

found in 80% of the dissected nerves, it was from

9.5–21 cm from the emergence of the intercostal

nerve in the posterior intercostal space. The bifurca-

tion was more proximal for the lower intercostal

spaces. The statistical analysis of our results does

not make it possible to describe this bifurcation per-

fectly. It would seem logical that this bifurcation is

further away from the nerve exit form the posterior

intercostals space in the largest individuals. We only

found this correlation with respect to the eleventh

intercostal nerve.

Neurotization of the brachial plexus is an estab-

lished procedure in the upper limb. Neurotization of

the lower limb remains experimental. Previous ana-

tomical studies have shown anatomical feasibility of

lumbar plexus neurotization using lower intercostals

nerves. These studies described a limited number of

cases and intercostal nerve harvesting was carried

out via an extensive surgical approach as described

by Lim and Baskaran (2001). Lang et al. (2004) have

reported recently their use of lower intercostal

nerves for neurotization of roots of the lumbosacral

plexus in humans. The overall length of nerve that

we succeeded in harvesting with an incision at the

lateral border of the latissimus dorsi varied from 10–

27 cm with an average value going 17.96 cm for the

ninth nerve intercostal to 15.94 cm for the eleventh

intercostal nerve. Even if the harvested nerve length

was not positively correlated with the trunk size in

our experiment, this length was largely sufficient to

carry out an anastomosis between the harvested

intercostal nerve and the roots at the level of cauda

equina. Only in three cases of the 300 harvesting

procedures (in which we had technical difficulties

because of especially thin nerves) we had sufficient

length to perform root neurotization.

This anatomical study of the ninth, tenth, and

eleventh intercostal nerves in the posterior and lat-

eral intercostal spaces confirms the feasibility of reli-

able intercostal nerve harvesting for neurotisation.

The bifurcation of the nerve into a deep and superfi-

cial branch remain incompletely defined but will

hopefully be refined with an increase in the number

of cases studied. The length of nerve harvested by

means of a second incision at the lateral border of

the latissimus dorsi muscle seemss sufficient to be

able to perform a nervous anastomosis at the level of

the conus medullaris. Potentially, selective neuroti-

zation could be carried out to lumbar roots supplying

important lower limb nerves such as the superior glu-teal nerve (for hip extension), the femoral nerve (for

knee extension), and the sciatic nerve (for knee flex-

ion and distal muscular functions). Clinical protocols

studying axonal re-growth after such neurotization

procedures using intercostals nerves are required.

We are currently conducting such an experimental

study in mammalian nerve tissue.

ACKNOWLEDGMENTS

This study profited from the financial support

of the Societe Francaise de Chirurgie Orthopedique

et Traumatologique and the Fondation pour la

Recherche Medicale.

REFERENCES

Asfazadourian H, Tramond B, Dauge MC, Oberlin C. 1999.

Morphometric study of the upper intercostal nerves: prac-

tical application for neurotisations in traumatic brachial

plexus palsies. Ann Chir Main 18:243–253.

Aubert J, Koumare K, Dufrenot A. 1981. [Anatomical study

of the twelfth intercostal nerve and oblique lumbotomies].

[in French] J Urol 87:283–289.

Davies F, Gladstone RJ, Stibbe EP. 1932. The anatomy of

the intercostal nerves. J Anat 66:323–333.

Hardy PAJ. 1988. Anatomical variation in the position of the

proximal intercostal nerve. Br J Anaesth 61:338–339.

Hauge EN. 1991. The anatomical basis of a new method for

re-innervation of the gluteal region in paraplegics. Acta

Physiol Scand 143:19–21.

Lang EM, Borges J, Carlstedt T. 2004. Surgical treatment of

lumbosacral plexus injuries. J Neurosurg Spine 1:64–71.

Lim HH, Baskaran . 2001. Intercostal nerve neurotization of

the femoral nerve in patients with paraplegia—an anatomi-

cal feasibility study. Med J Malaysia 56:61–65.

Rouviere H. 1967. Anatomie humaine descriptive et topogra-

phique, Tome II. Paris: Tronc.

57Anatomical Study of the Lower Intercostal Nerves

RETRACTED

Sakamoto H, Akita K, Sato T. 1996a. An anatomical analysis

of the relationships between the intercostal nerves and the

thoracic and abdominal muscles in man. I. Ramification of

the intercostal nerves. Acta Anat 156:132–142.

Sakamoto H, Akita K, Sato T. 1996b. An anatomical analysis

of the relationships between the intercostal nerves and the

thoracic and abdominal muscles in man. II. Detailed anal-

ysis of innervation of the three lateral abdominal muscles.

Acta Anat 156:143–150.

Schalow G, Aho A, Lang G. 1992. Microanatomy and number

of nerve fibres of the lower intercostal nerves with respect

to a nerve anastomosis. Donor nerve analysis. I. (IV). Elec-

tromyogr Clin Neurophysiol 32:171–185.

Tok S, Schmid UD, Ferbert A, Davenport T. 1991. Inter

costolumbar spinal nerve anastomosis. An experimental

study in dogs. Spine 16:463–466.

Walmsley T. 1916. The costal musculature. J Anat 50:165–171.

Williams P, Warwick R, Dyson M, Bannister L. (eds.) 1989.

Gray’s anatomy. 37 Ed. Edinburgh: Lippincott.

Yan L, Zhong S. 2000. [Applied anatomy of the transverse

rectus abdominis musculocutaneous (TRAM) flap in rela-

tion to the eleventh or twelfth thoracic nerve]. Zhonghua

Zheng Xing Wai Ke Za Zhi 16:81–83.

Zhao S, Beuerman RW, Kline DG. 1997. Neurotization of

motor nerves innervating the lower extremity by utilizing

the lower intercostal nerves. J Reconstr Microsurg 13:39–45.

58 Vialle et al.