world neurosurg 2011_ p170.pdf

Tptie

PEER-REVIEW REPORTS

C2 Nerve Root Sectioning in Posterior C1-2 Instrumented Fusions
Matthew M. Kang, Erich G. Anderer, Robert E. Elliott, Stephen P. Kalhorn, Anthony Frempong-Boadu
M

Ttsp

SDuaptCmmf

AantPtcsaCdwstast

imgaca

�

uh2dsp

�atdmc

INTRODUCTION

The clinical consequences of sacrificingthe C2 nerve root when performing poste-rior atlantoaxial instrumentation is notwell described in the literature (1, 6, 27).

wo of the largest series describing C1-2osterior instrumented fusions reported

hat section of the C2 nerve root eased thensertion of C1 lateral mass screws, andxcellent clinical outcomes were seen (1,

6). However, neither of the two studiessystematically analyzed the clinical con-sequences of C2 nerve transection. Ana-lyzing outcomes after C1-2 posterior in-strumented fusion with polyaxial screwand rod constructs in 23 elderly patients,Squires and Molinari (27) reportedshorter operative times, decreased bloodloss, and minimal clinical consequenceswhen the C2 nerve was sectioned. We re-port our experience in 20 consecutive pa-tients in which the C2 nerve root was sec-tioned during posterior atlantoaxialfusion. Specifically, we discuss the tech-nique of C2 nerve transection, its per-

Key words� Atlantoaxial� C1-2 instability� C2 ganglion� Dens fracture� Neuropathic pain

Abbreviations and AcronymsCSF: Cerebrospinal fluidCT: computed tomographyVAS: Visual analogue scale

Department of Neurosurgery, New YorkUniversity Langone Medical Center, Bellevue

Hospital, New York, New York, USA

To whom correspondence should be addressed:Stephen Kalhorn, M.D. [E-mail: [email protected]]

Citation: World Neurosurg. (2012) 78, 1/2:170-177.DOI: 10.1016/j.wneu.2011.07.010

Journal homepage: www.WORLDNEUROSURGERY.org

Available online: www.sciencedirect.com

1878-8750/$ - see front matter © 2012 Elsevier Inc.All rights reserved.

ceived advantages, and clinical outcomes. a

170 www.SCIENCEDIRECT.com

ETHODS

he institutional review board at our insti-ution approved this study, and verbal con-ent was obtained from all patients beforearticipating in the study.

urgical Techniqueuring the years 2002–2010, 20 patientsnderwent posterolateral instrumentationnd fusion using C1 lateral mass and C2ars interarticularis or pedicle screws by

wo neurosurgeons at our institution (Paulooper, M.D., and A.F.B). Preoperativeagnetic resonance imaging, computed to-ography (CT), and radiographs were per-

ormed to assess pertinent neurovascular

� OBJECTIVE: To analyze qualitativelyin C1-2 posterolateral instrumented fu

� METHODS: From 2002–2010, 20 consnstrumented fusions using C1 latera

ainly for type 2 dens fractures. Screwuidance using standard techniques.nd transected in all patients. Mean folinical and radiographic examinationnalogue scale (VAS) questionnaires

RESULTS: One patient was lost tovisit. Fusion was evident in all patientthree patients with <12 months of

nintended neurologic deficits, vasculardware malfunction or malposition.0 patients complained of sensory dististribution. After longer follow-up,ensory symptoms. Quality of life wasatient developed neuropathic pain at

CONCLUSIONS: This study is the fifter posterior atlantoaxial instrumente

he C2 nerve root increases fusion suecortication of the atlantoaxial jointent, and decreases blood loss an

onsequences.

nd osseous anatomy and spinal stability. t

WORLD NEUROSURGE

ll patients were counseled before surgerybout the occurrence of postoperativeumbness or neuropathic pain and the ra-

ionale for transection of the C2 nerve root.atients were placed prone in a Mayfield

hree-point head holder and Miami-J hardollar and positioned prone on gel rolls on atandard operating table. Lateral intraoper-tive fluoroscopy was used to visualize the1-2 level and assess alignment. Closed re-uction was attempted if such a maneuveras necessary and could be performed

afely. The head holder was securely at-ached to the operating table to maintainlignment. A standard midline incision andubperiosteal dissection with electrocau-ery provided exposure from the suboccipi-

nerve dysfunction after its transections.

ive patients underwent posterolateralss and C2 pars or pedicle screws,ere placed under lateral fluoroscopiceral C2 nerve roots were coagulated-up was 30.7 months and consisted ofephone interviews, and mailed visualssing C2 nerve dysfunction.

ow-up after the initial postoperativeth 12 months of follow-up and two ofw-up. There were no instances of

jury, cerebrospinal fluid (CSF) leak, orhe 2-week or 6-week office visit, 4 ofnce, and 2 had paresthesias in the C2

additional patient developed mildrsely affected in only one patient. Notime after C2 sectioning.

series to describe C2 nerve functionsion in adults of all ages. Sacrifice ofe, allows for better preparation androves visualization for screw place-erative time without major clinical

C2sion

ecutl mas w

Bilatllows, telasse

folls wifollo

ar inBy turbaone

adveany

rstd furfac

, impd op

al area to the C2-3 joint.

RY, DOI:10.1016/j.wneu.2011.07.010

mailto:[email protected]

http://www.worldneurosurgery.org

http://www.sciencedirect.com

2dipmlLjmi

szDvllarw

Cn

wcat8ahdatlcowisdpPgo

CoeiAdiQ(itcirrdraa

PPwagdtwamdma

aa

pcnpdiicslnstqs(Tpqspqtspp

STEeacswa

R

PFTmtomTsmcht

PEER-REVIEW REPORTS

MATTHEW M. KANG ET AL. C2 NERVE SECTION IN C1-2 POSTEROLATERAL FUSION

As the C1-2 joint is approached, the C2nerve root and its dense epidural venousplexus are encountered (20, 24, 25, 28).Stepwise coagulation and cutting of the C2nerve root, surrounding veins, and the fi-brous sheath are performed, from dorsal toventral, using bipolar electrocautery (Figure1). The C2 ganglion is located within theconfines of the C2 intervertebral “fora-men,” which is bordered superiorly by theposterior arch of the atlas, inferiorly by thelamina of the axis, and anteriorly by the at-lantoaxial joint. Posteriorly, the “foramen”is not enclosed by bone but rather a fibroussheath contiguous with the ligamentumcalled the atlantoepistrophic ligament (20, 24,

8). The C2 nerve root continues almostirectly horizontally from the ganglion, and

t is transected directly posterior to the mid-oint of the atlantoaxial joint and C1 lateralass to avoid cerebrospinal fluid (CSF)

eak. For “distal-type” ganglia as defined byu and Ebraheim (20), we incise the nerve

ust proximal to the ganglion itself. For theore common “proximal-type” ganglia, we

ncise the ganglion itself.The technique for nerve root section con-

ists of progressive bipolar electrocauteri-ation and incision with Malis scissors.eep to the nerve, the ventral portion of the

enous plexus is encountered and coagu-ated, exposing the posterior cortex of theateral mass of C1 just below the posteriorrch. The use of surgical clips or ties on theemaining nerve root stump is optional andas not performed in our series.Under fluoroscopy, C1 lateral mass and

2 pars screws are placed using the tech-ique described by Harms and Melcher (12)

and connected using rods. Later in our ex-

Figure 1. Illustration showing the placementof a C1 lateral mass screw on the right sideafter sectioning the C2 nerve root on thatside and coagulating the large venous plexusthat typically resides in this region.

perience, we have used C2 pedicle screws o

WORLD NEUROSURGERY 78 [1/2]: 170-1

hen the anatomy is appropriate (ie, pedi-le of adequate diameter, normal calibernd location of the vertebral artery). Similaro the modifications described by others (1,), we insert a small curette into the atlanto-xial joint to remove the synovial lining. Aigh-speed matchstick burr is used toecorticate the articular surfaces within thetlantoaxial joint and the dorsal surfaces ofhe C1 and 2 lateral masses, C2 pars andamina, and C1 posterior arch. Morselized,ancellous cadaveric bone graft is placedver the decorticated bony surfaces andithin the C1-2 joint (Figure 2). The wound

s closed in multiple layers with absorbableutures and a drain is typically not placed. Ary dressing is applied, and a soft collar islaced and used as needed for comfort.ostoperative CT scan or standard radio-raphs are obtained on postoperative day 1r 2 to assess hardware position.

Three patients had minimally invasive1-2 fusions; this technique has been previ-usly described (13). It entails small bilat-ral incisions, centered over C2 and approx-mately 2 cm in length and 2 cm off midline.fter the cervical fascia is opened, tubularilators are inserted and docked on the lam-

na of C2. After serial dilators are inserted, auadrant expandable tubular retractor

Medtronic, Memphis, Tennessee, USA) isnserted. Following expansion of the retrac-or blades, a subperiosteal dissection isompleted with monopolar cautery provid-ng excellent exposure from the C1 poste-ior arch down to the C2-3 facet joint. Theemaining steps of nerve root sacrifice, jointecortication and packing, and screw andod insertion are the same as in the openpproach. The wound is closed in layers,nd no drain is placed.

ostoperative Follow-up and Assessmentatients typically had office visits at 2eeks; 6 weeks; and at 3, 6, and 12 months

fter surgery. The initial 2-week visit was foreneral evaluation and wound check. Ra-iographs of the cervical spine (anteropos-

erior, lateral, flexion-extension views)ere obtained at 3 months and at 1 year to

ssess for stability of the construct, align-ent, and evidence of fusion. Fusion was

eemed successful based on lack of move-ent on flexion-extension radiographs

cross the C1-2 joint or contiguous bridging
f trabecular bone across the C1-2 level as n
77, JULY/AUGUST 2012 ww

ssessed on CT usually performed 1 yearfter surgery.

Follow-up consisted of office visits, tele-hone interviews, and completion of a spe-ialized visual analog scale (VAS) question-aire. The attending neurosurgeon (A.F.B.)erformed all neurologic examinationsuring the clinic visits. Formal sensory test-

ng was not performed on all patients earlyn the series. The C2-specific questionnaireonsisted of a series of questions that as-essed the incidence and severity of radicu-ar pain, dysesthesias, paresthesias, andumbness (Table 1). To assess how bother-ome each patient found his or her symp-oms, the severity of each parameter wasuantified using the 10-point VAS. Thecores ranged from 0 (asymptomatic) to 10symptomatic and severely bothersome).o limit the impact of surgeon bias, a nurseractitioner performed all the phone calluestionnaires and tabulated the written re-ponses that were returned via mail. Eightatients were queried again using the VASuestionnaire approximately 2 years after

he initial evaluation to assess change inymptoms over time. Patients were not ap-rised of their prior VAS values before re-eat query.

tatistical Analysishe raw data were entered into Microsoftxcel (Office 2008 for Mac). Averages arexpressed as means and medians. Preoper-tive and postoperative VAS values wereompared using paired-samples Wilcoxonigned rank test. A two-tailed P value � 0.05as considered statistically significant for

ll analyses.

ESULTS

atient Demographics, Presentation, andollow-upable 2 summarizes the preoperative de-ographic and clinical data and postopera-

ive outcomes. The study cohort consistedf 11 women (55%) and 9 men (45%) with aean age of 66 years (range 19 – 89 years).

he most common indication for C1-2 fu-ion was type 2 dens fracture (65%), and theost common symptom was axial (me-

hanical) neck pain (70%). Three patientsad preoperative dysesthetic pain and pares-

hesias. No patient had preoperative numb-
ess on examination.
w.WORLDNEUROSURGERY.org 171

PEER-REVIEW REPORTS


172

Figure 2. This 83-year-old woman presented with neck pain and quadriparesis after a fall. Computed tomography (CT) scan showed atype 2 dens fracture (A) with retropulsion of the fragment into the canal causing compression of the cervicomedullary junction. She wasplaced in traction that provided nearly complete reduction and realignment of C2 (B). She then underwent C1-2 posterolateralinstrumentation and fusion with bilateral sectioning of the C2 nerve root. Sagittal CT images show proper placement of the C2 pediclescrews, bicortical purchase of both C1 lateral mass screws, and allograft within the atlantoaxial joint (C and D). Axial images confirmproper placement of C1 (E) and C2 (F) screws in relation to the vertebral arteries. Fusion across the C1-2 joint was evident on the 6-
month follow-up CT scan (G–I).
www.SCIENCEDIRECT.com WORLD NEUROSURGERY, DOI:10.1016/j.wneu.2011.07.010

po

CAwtatthgTtottlsp

a(ld

Tim2tclfinrN(phbtaaw

(1hwttcintodnP0

OTthdosmtfAsi

D

Tan

lity o

PEER-REVIEW REPORTS


All C1 screws were placed into the lateralmasses. Bilateral screws were placed intothe pars interarticularis in 15 patients(75%); 3 of these patients had minimallyinvasive surgeries. Three patients had bilat-eral C2 pedicle screws, and two patients hadpars screws placed ipsilaterally to a high-riding vertebral artery and contralateralpedicle screws. Mean estimated blood lossfor the entire cohort was 194 mL (range 50 –500 mL).

The mean follow-up for the entire cohortwas 30.7 months (median 32 months,range 7–72 months). One patient was lostto follow-up after the initial 2-week postop-erative visit and could not be reached. Hard-ware was appropriately positioned in all pa-tients on postoperative imaging. One C2pars screw violated the vertebral artery fora-men without consequence. No patient hadevidence of instrumentation failure or non-union based on postoperative imaging. Fu-sion was evident in all patients with 12months of follow-up and two of three pa-tients with �12 months of follow-up. Post-operative axial neck pain resolved or im-

Table 1. C2 Nerve Root Dysfunction QueScale

Numbness

1. Do you have numbness in the occipital region

2. On a scale of 1–10, how severe is the numb

3. How often do these symptoms bother you?

4. Does this symptom negatively affect your qua

Paresthesias

1. Do you have paresthesias in the occipital reg

2. On a scale of 1–10, how severe are the pare



Radicular Pain (Shooting Pain)

1. Do you have radicular pain in the occipital re

2. On a scale of 1–10, how severe is the radicu



Dysesthesia (Burning Discomfort or Pain)

1. Do you have burning pain or discomfort in th

2. On a scale of 1–10, how severe is the burnin



proved in all patients. All three patients with c


reoperative dysesthesia had complete res-lution of this pain after stabilization.

2 Nerve Root Dysfunctiont the initial office visits at 2 weeks and 6eeks after surgery, only 4 (20%) of 20 pa-

ients complained of occipital numbness,nd 2 (10%) of these patients had C2 pares-hesias. On examination, five (25%) pa-ients had occipital anesthesia, and 5 othersad hypesthesia but were able to distin-uish laterality and presence of a stimulus.hree patients (one with anesthesia and

wo with hypesthesia) became awaref their deficits only after testing. No pa-

ients had radicular or dysesthetic pain athis visit or at any time during the fol-ow-up period. No patient had speech orwallowing difficulties or ear pain at anyoint during follow-up.

One patient was lost to further follow-upnd is not included in the outcome analysesn � 19). The remaining patients had ateast 6 months of postsurgical follow-up. Inelayed follow-up, a fifth patient (26.3%)

aire Assessed by Visual Analogue

Yes or no

1–10

Daily, weekly, rarely

f life? Yes or no

Yes or no

as? 1–10


f life? Yes or no

Yes or no

in? 1–10


f life? Yes or no

pital region? Yes or no

or discomfort? 1–10


f life? Yes or no

omplained of symptomatic numbness. p


he VAS questionnaire was initially admin-stered to 17 (89.5%) of 19 patients at a

ean follow-up time of 21 months (median2 months, range 4 –58 months). Three pa-ients did not return the survey by mail andould not be reached by phone. Their fol-ow-up was censored at the time of last of-ce visit. The mean VAS scores for numb-ess and paresthesias were 1.2 (median 1,ange 0 –5) and 0.4 (median 0, range 0 –3).o patient had radicular or dysesthetic pain

VAS score 0). Only 1 (5.3%) patient com-lained that numbness adversely affectedis quality of life and affected him on a dailyasis (patient 5, VAS score 5). No other pa-

ients reported that their C2 symptoms neg-tively affected their overall quality of life,nd the frequency of occurrence was on aeekly to monthly basis.At a mean of 37.6 months after surgery

median 45 months, range 7–72 months),1 patients were queried again. No patientad worsened or new symptoms, and thereas no clinical evidence of neuroma forma-

ion. Six patients with numbness or pares-hesias rated their symptoms as stable (nohange in VAS rating). Two patients hadmprovement in their paresthesias orumbness as measured by VAS. Three pa-

ients continued to be asymptomatic with-ut sensory loss. There were no significantifferences in the early vs late VAS values forumbness (mean/median 1.55/1 vs 1.36/1;� 0.16) or paresthesias (mean/median

.64/0 vs 0.45/0; P � 0.16).

perative Complicationshere were no instances of neurologic de-

erioration, vascular injury, CSF leak, orardware failure. One patient developed aeep wound infection requiring removalf instrumentation, evacuation of ab-cess, intravenous antibiotics, and place-ent of a halo vest for 6 months. Al-

hough the patient was a heavy smoker,usion eventually occurred in this patient.nother patient developed a superficialubcutaneous infection that resolved withntravenous antibiotics.

ISCUSSION

he technique of C2 nerve root sacrifice toid the insertion of C1 instrumentation isot new (1, 6, 8, 13, 27). However, there is a

stionn

?

ness?

lity o

ion?

sthesi

lity o

gion?

lar pa

lity o

e occi

g pain

aucity of data on the incidence and severity


Table 2. Preoperative Demographic and Clinical Data and Postoperative Outcomes for 20 Consecutive Patients Who Underwent C1-2 Posterolateral Instrumentationand Fusion with Sectioning of the C2 Nerve Root

PatientAge

(years) Gender Diagnosis Preoperative Symptoms Surgical Technique EBL (mL)PostoperativeC2 Symptoms

NumbnessVAS

ParesthesiasVAS

Follow-upDuration(months)

1 75 F Dens fracture (acute) Neck pain C1 lateral mass/C2 pars 500 Hypesthesia 1 0 45

2 81 F C1-2 instability, synovial cyst Myelopathy/weakness C1 lateral mass/C2 pars 270 Anesthesia 3 2 50

3 65 M Dens fracture (acute) Neck pain C1 lateral mass/C2 pars (MIS) 50 Hypesthesia 1 0 34

4 83 F RA Myelopathy/weakness C1 lateral mass/C2 pars 250 None 0 0 25

5 74 M Dens fracture (acute) Neck pain C1 lateral mass/C2 pars 50 Anesthesia 5 3 36

6 58 F RA Neck pain C1 lateral mass/C2 pars 300 Anesthesia 2 0 46

7 79 F Dens fracture (chronic nonunion) Neck pain C1 lateral mass/C2 pars 300 None NA NA 18

8 89 F Dens fracture (failed odontoidscrew)

Neck pain C1 lateral mass/C2 pars 100 None 0 0 35

9 79 M Dens fracture (acute), occipitalneuralgia

Occipital/neck pain C1 lateral mass/C2 pars 150 None 0 0 72

10 43 M Dens fracture (chronic nonunion) Neck pain C1 lateral mass/C2 pars (MIS) 100 Hypesthesia 2 1 48.5

11 37 F Os odontoideum Neck pain C1 lateral mass/C2 pars (MIS) 200 Hypesthesia 1 0 32

12 54 M Dens fracture (acute) Neck pain C1 lateral mass/C2 pars 100 NA NA NA 0.5

13 80 F Dens fracture (acute) Neck pain C1 lateral mass/C2 pars 400 None NA NA 25

14 76 F C1-2 instability, occipitalneuralgia

Neck/occipital/jaw pain C1 lateral mass/C2 pars 150 None 1 0 22

15 19 M Os odontoideum Neck pain C1 lateral mass/C2 pedicle 150 Anesthesia 2 0 21

16 76 F C1-2 instability, occipitalneuralgia

Neck/occipital/jaw pain C1 lateral mass/C2 pars 50 None 0 0 39

17 75 M Dens fracture, Jefferson burstfracture

Neck pain C1 lateral mass/C2 pedicleand pars*

100 None 0 0 13

18 40 M Dens fracture, Jefferson burstfracture

Neck pain C1 lateral mass/C2 pedicleand pars*

200 Anesthesia 2 1 8

19 83 F Dens fracture (acute) Central cord syndrome C1 lateral mass/C2 pedicle 200 None 0 0 7

20 57 M Dens fracture (acute) Neck pain C1 lateral mass/C2 pedicle 150 Hypesthesia 1 0 7.5

EBL, estimated blood loss; MIS, minimally invasive spine surgery; NA, not available; RA, rheumatoid arthritis; VAS, visual analogue scale.*A high-riding vertebral artery was present on the one side; therefore, a pars screw was placed ipsilaterally to the anomalous artery and a pedicle screw contralaterally.

PEER-R

EVIEWR

EPORTS

MATTH

EW

M.KAN

GET

AL.

C2

NER

VE

SEC

TIO

NIN

C1-2

PO

STER

OLATER

AL

FU

SIO

N

174

ww

w.S

CIE

NC

ED

IREC

T.c

om

WO

RLD

NEU

RO

SU

RG

ER

Y,

DO

I:10.1

016/j

.wneu.2

01

1.0

7.0

10

hqtsqrc

novOpnn(optgtdosninrWppCt

drwinppdapttwogftfot

g

ocsmunwqt2snarcpItirns

mtsSdpnmsrwtsetociCh

rAstCirgiedo

PEER-REVIEW REPORTS


of C2 nerve dysfunction after its transec-tion. Initially reported as a useful techniquein 1994 (8), Goel et al. (6) later reported theadvantages of C2 nerve sectioning in a se-ries of 160 patients with C1-2 posterior fu-sions using screw and rigid place fixation.They reported that sectioning the C2 nerveroot improved exposure for screw place-ment and the surface area of fusion. C2nerve dysfunction was not specifically elic-ited as part of the surgical follow-up, but 18of 157 patients with adequate follow-upcomplained of sensory disturbances in theC2 distribution (11.5%; mean follow-up 42months). None of these disturbances wereclinically significant or bothersome to thepatients, and there were no instances ofdysesthesia. Similarly, Aryan et al. (1) re-ported their success with a modified Harmstechnique with transection of the C2 nervein 102 patients who underwent C1-2 poste-rior fixation with polyaxial screw and rodconstructs. They reported postoperativeneuropathic pain in a single patient (1%)but did not report on the incidence of hyp-esthesia.

To our knowledge, the only direct exam-ination of the clinical consequences of C2nerve root sacrifice for posterior atlantoax-ial fusions was performed by Squires andMolinari (27). In a series of 23 elderly pa-tients who underwent C1-2 fixation withpolyaxial screw and rod constructs, theseauthors retrospectively compared the out-comes of 18 patients who had C2 nerve rootsacrifice and 5 patients who did not have C2nerve root sacrifice. They reported trendstoward shorter operative times (mean 109minutes vs 187 minutes) and less blood loss(mean 344 mL vs 1030 mL) with C2 nervesacrifice. A single patient complained ofbothersome numbness (4.5%), but no in-stances of neuropathic pain occurred.Given such limited clinical consequences,Squires and Molinari (27) recommendedconsideration of C2 nerve sectioning in el-derly patients to limit the blood loss andoperative times in these vulnerable patients.

In our series of 20 adults who had C1-2instrumented fusions with polyaxial screwand rod constructs and bilateral C2 nervesacrifice, we noted absence of C2 symptomsin half of patients. Five (25%) patients hadC2 nerve distribution paresthesias or both-ersome anesthesia, but these symptoms didnot interfere significantly with their every-
day lives. Five further patients had minimal e

ypesthesia that was not bothersome. VASuestionnaires revealed low scores across

he domains of numbness and paresthe-ias, indicating minimal impact on theiruality of life. There were no cases of neu-opathic or radicular pain, and no delayedases of neuroma formation.

Excluding one patient lost to follow-up,ine of our patients denied any numbnessn direct examination, telephone inter-iews, or returned mail questionnaires.ther authors have presented similar re-orts stating that transection of the C2erve did not lead to clinically significantumbness or dysesthesia in their patients

1, 6, 27). It has been postulated that relieff preoperative symptoms (i.e., axial neckain, dysesthesia, radicular pain) in a pa-

ient after successful fusion may havereatly diminished the perception of anes-

hesia (6). This lack of noticeable sensoryisturbance may also be due to the overlapf dermatomes in the human body. Expan-ion and contraction of dermatomes fromeuroplasticity after central nervous system

njury or local ingrowth after peripheralerve or nerve root damage or schwannomaesection have been reported (15, 16, 26).

hatever the explanation may be, at leastartial greater occipital sensation could bereserved with surgical transection of both2 nerve roots, and the potential for adap-

ation over time may exist.In our series, three of the five patients who

eveloped bothersome symptoms did so in aelatively short time after surgery (2–6eeks). It took a longer time for the remain-

ng 2 patients to develop appreciable numb-ess or paresthesias (4–6 months). Based onatient interviews, this delayed awareness ofaresthesias was possibly due to the subsi-ence of postsurgical pain leading to a greaterwareness of such paresthesias (unmaskinghenomenon). Studying patients with occipi-

al neuralgia, Lozano et al. (19) noted that pa-ients who failed ganglionectomy presentedith immediate pain and did not have a periodf relief followed by pain. This finding sug-ests that the lack of noticeable C2 nerve dys-unction in postoperative patients may reflecthe low long-term likelihood of forming pain-ul neuralgias. We observed this in our seriesf patients on extended follow-up with no pa-

ients developing delayed pain or neuromas.Incomplete destruction of the dorsal root

anglion may result in pain secondary to ab-
rrant axonal regeneration or the formation t

f neuromas (31). Patients who undergohemical neurolysis or simple decompres-ion (opening of the atlantoepistrophic liga-ent) are more likely to have functional resid-

al nerve root pain from the formation ofeuromas or residual root irritation; this painas relieved in most patients, who subse-uently underwent complete surgical resec-

ion of the C2 nerve root and ganglion (19,8). Many centers recommend complete exci-ion of the root ganglion in cases of occipitaleuralgia. Because the dorsal root ganglionrises approximately 2–4 mm distal to the du-al ring of C2 (17, 31), the root excision can bearried toward the dural ring but not moreroximally to minimize the risk of CSF leak.

n our series, the ganglion itself was oftenransected (leaving a portion of the ganglionn situ). However, no patient developed neu-opathic pain over time, supporting the tech-ique of transection at the point of screw in-ertion to limit the risk of CSF leak.

Table 3 summarizes the outcomes of theajor studies reporting on posterior atlan-

oaxial fixation with and without C2 nerveacrifice (1-3, 6, 11, 12, 18, 21, 23, 27, 29).ectioning seems to result in a higher inci-ence of hypesthesia, whereas rates of re-orted neuropathic pain are higher witherve root preservation and screw place-ent via caudal retraction. The rates of fu-

ion in the major series are remarkably highegardless of C2 nerve technique. Thereere, however, trends of shorter operative

imes and less blood loss with nerve rootacrifice. Some centers have reported theirxperience using partially threaded screwshat have smooth shanks to limit irritationf the nerve (11). Other centers have advo-ated an alternative starting point for screwnsertion directly on the posterior arch of1 (21, 30). However, C2 nerve dysfunctionas been reported with both techniques.

Intraoperatively, sectioning the C2 nerveoot offered several technical advantages.s shown in Figure 1, the C2 ganglion andurrounding venous plexus occupy most ofhe space between the C1 posterior arch and2 lamina. The exposure gained by section-

ng the C2 nerve root can provide ampleoom for the proper caudad to cephalad an-ulation of the C1 screw without traction

njury to the C2 nerve root. The increasedxposure can also eliminate the need forrilling the posterior arch of C1, a maneuverccasionally needed to gain the proper ver-

ical angle for screw placement and one that


twnrcMofl

-surgica

PEER-REVIEW REPORTS


puts the vertebral artery at risk. Completeexposure of the atlas allows for improvedidentification of anomalous vertebral arter-ies (14) and optimal entry point and angula-tion for the proper placement of C2 pediclescrews (4, 6, 8, 9).

A further and, in our opinion, critical ad-vantage of nerve sacrifice is the excellentexposure it provides of the atlantoaxialjoints. Originally described by Goel and La-heri in 1994 (8), arthrodesis can be im-proved with C1-2 synovectomy and packingof the joint with bone graft. A more recentadaptation of this technique involves the in-sertion of titanium spacers filled with bonegraft into the decorticated atlantoaxial joint(4 –7, 10). In vitro biomechanical studieshave shown the addition of such spacersprovides more construct rigidity comparedwith C1-2 instrumentation alone (22). Suchthorough synovectomy, joint surface decor-tications, and bone graft or spacer insertionis more difficult if the C2 nerve remains

Table 3. Reported Rates of C2 Neuralgia

Study and TechniqueNumber of

Patients

C2 Nerve Preservation

Harms and Melcher, 2001 (12) 37

Gunnarsson et al., 2007 (11) 25

Stulik et al., 2007 (29) 28

De Iure et al., 2009 (3) 14

Payer et al., 2009 (23) 11*

Conroy et al., 2010 (2) 9

Lee et al, 2010 (18) 27

Pan et al., 2010 (21) 48

Squires and Molinari, 2010 (27) 5

C2 Nerve Section

Goel et al., 2002 (6) 160§

Aryan et al., 2008 (1) 102

Squires and Molinari, 2010 (27) 18

Current study 20

EBL, estimated blood loss.*The study consisted of 12 patients; a single patient und†Neuralgia resolved in one patient by 6 weeks after surg‡All instances of postoperative numbness occurred in patie

No instances occurred since adoption of the entrance§Three patients died in the perioperative period.�One patient was lost to follow-up after the 2-week post

intact.

176 www.SCIENCEDIRECT.com

Stepwise dorsal to ventral coagulation ofhe C2 nerve reduces bleeding from theell-described venous plexus around theerve. Minimal trauma to this plexus canesult in profuse bleeding that can be diffi-ult to control without damage to the nerve.odern studies reporting mean estimates

f blood loss in C1-2 posterolateral fusionsrequently describe large volumes of bloodoss of �1 L (18, 21, 27, 29). Our average

estimated blood loss was approximately200 mL with several cases of �100 mL, sig-nificantly less than that reported by manyother centers with C2 nerve preservation.

Study LimitationsThe major limitations of our study includeits small sample size, lack of control group,and retrospective methodology. A furtherlimitation is the lack of documentation ofthe preoperative and postoperative painmedications or dosages; we were unable to

ysfunction After C1 Lateral Mass Screw P

BL Mean(mL) Range (mL)

Mean OperativeTime (minutes)

— — —

— — —

540 50–1500 83

— — —

480 150–800 155

— — —

730 100–1500 169

— — —

1030 200–1800 187

— — —

— — —

344 50–750 109

194 50–500 —

C2 nerve sectioning without adverse effect.persisted in two other patients.o underwent standard C1 lateral mass screw insertion via thscrew on the posterior arch of C1.

l office visit.

determine narcotic or analgesic require-

WORLD NEUROSURGE

ments. Although patients were counseledpreoperatively regarding the expected se-quelae of C2 nerve root transaction, andpostoperative office visits assessed deficitsand symptoms, this was not done quantita-tively. Formal sensory testing was not per-formed on all patients. We did not quantifythe change in the area of hypesthesia overtime; this was assessed subjectively bypatient query. The detailed C2 nerve dys-function questionnaire was not adminis-tered preoperatively to establish a baselinefor comparison and has not been validatedby larger studies. The specific C2 nervequestionnaire was created and adminis-tered in a delayed fashion, prohibiting di-rect comparison from the immediate post-operative period. To mitigate this deficiencyand to assess change over time, we queriednine patients again and reassessed their lev-els of delayed or persistent C2 nerve dysfunc-tion. These delayed evaluations showed thatsome patients became accustomed to their

ent

Pain orysesthesia Symptomatic Numbness

mber % Number %

0 0 0 0

3 12 0 0

3 10.7 0 0

0 0 0 0

3 27.3 0 0

3† 33.3 0 0

1 3.7 0 0

0 0 4‡ 8.3

0 0 0 0

0 0 18 11.5

1 1.0 — —

0 0 1 4.5

0 0 5 26.3�

niques described by Harms and Melcher (4 of 12 [33.3%]).

or D lacem

ED

Nu

erwentery butnts wh e techof the

deficits over time without an increase in

RY, DOI:10.1016/j.wneu.2011.07.010

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

3

3

Cacc

Rp

CD

J

A

PEER-REVIEW REPORTS


symptoms or the development of neuropathicpain or neuromas. Prospective, randomizedtrials using validated patient satisfaction met-rics are essential to compare these techniquesdirectly and determine the optimal methodsof C1 lateral mass screw placement.

CONCLUSIONS

Sectioning of the C2 nerve root in posterioratlantoaxial instrumented fusion is contro-versial and not standard practice. Althoughlarger series have been performed, we pro-vide one of the first reports describing thepresence and severity of C2 dysfunction af-ter C2 nerve sacrifice. Our review is limitedby its retrospective nature, small cohortsize, and simple clinical grading scale.Based on our experience, we are hesitant topreserve the C2 nerve root when placing C1lateral mass screws. A review and follow-upof C2 dysfunction in patients from otherinstitutions, with or without C2 nerve tran-section, would help determine the idealmethod for placing C1 screws.

REFERENCES

1. Aryan HE, Newman CB, Nottmeier EW, Acosta FL Jr,Wang VY, Ames CP: Stabilization of the atlantoaxialcomplex via C-1 lateral mass and C-2 pedicle screwfixation in a multicenter clinical experience in 102patients: modification of the Harms and Goel tech-niques. J Neurosurg Spine 8:222-229, 2008.

2. Conroy E, Laing A, Kenneally R, Poynton AR: C1lateral mass screw-induced occipital neuralgia: a re-port of two cases. Eur Spine J 19:474-476, 2010.

3. De Iure F, Donthineni R, Boriani S: Outcomes of C1and C2 posterior screw fixation for upper cervicalspine fusion. Eur Spine J 18(Suppl 1):2-6, 2009.

4. Goel A: Treatment of basilar invagination by atlan-toaxial joint distraction and direct lateral mass fixa-tion. J Neurosurg Spine 1:281-286, 2004.

5. Goel A: Atlantoaxial joint jamming as a treatmentfor atlantoaxial dislocation: a preliminary report.Technical note. J Neurosurg Spine 7:90-94, 2007.

6. Goel A, Desai KI, Muzumdar DP: Atlantoaxial fixa-


7. Goel A, Kulkarni AG, Sharma P: Reduction of fixedatlantoaxial dislocation in 24 cases: technical note. JNeurosurg Spine 2:505-509, 2005.

8. Goel A, Laheri V: Plate and screw fixation for at-lanto-axial subluxation. Acta Neurochir (Wien) 129:47-53, 1994.

9. Goel A, Shah A: Atlantoaxial facet locking: treat-ment by facet manipulation and fixation: experiencein 14 cases. J Neurosurg Spine 14:3-9, 2011.

0. Goel A, Shah A, Gupta SR: Craniovertebral instabil-ity due to degenerative osteoarthritis of the atlanto-axial joints: analysis of the management of 108cases. J Neurosurg Spine 12:592-601, 2010.

1. Gunnarsson T, Massicotte EM, Govender PV, RajaRampersaud Y, Fehlings MG: The use of C1 lateralmass screws in complex cervical spine surgery: indi-cations, techniques, and outcome in a prospectiveconsecutive series of 25 cases. J Spinal Disord Tech20:308-316, 2007.

2. Harms J, Melcher RP: Posterior C1-C2 fusion withpolyaxial screw and rod fixation. Spine (Phila Pa1976) 26:2467-2471, 2001.

3. Holly LT, Isaacs RE, Frempong-Boadu AK: Mini-mally invasive atlantoaxial fusion. Neurosurgery 66:193-197, 2010.

4. Hong JT, Lee SW, Son BC, Sung JH, Kim IS, Park CK:Hypoglossal nerve palsy after posterior screw place-ment on the C-1 lateral mass: case report. J Neuro-surg Spine 5:83-85, 2006.

5. Htut M, Misra P, Anand P, Birch R, Carlstedt T: Painphenomena and sensory recovery following brachialplexus avulsion injury and surgical repairs. J HandSurg Br 31:596-605, 2006.

6. Kim P, Ebersold MJ, Onofrio BM, Quast LM: Surgeryof spinal nerve schwannoma: risk of neurologicaldeficit after resection of involved root. J Neurosurg71:810-814, 1989.

7. Kyoshima K, Uehara T, Koyama J, Idomari K, YomoS: Dumbbell C2 schwannomas involving both sen-sory and motor rootlets: report of two cases. Neuro-surgery 53:436-439, 2003.

8. Lee SH, Kim ES, Sung JK, Park YM, Eoh W: Clinicaland radiological comparison of treatment of atlan-toaxial instability by posterior C1-C2 transarticularscrew fixation or C1 lateral mass-C2 pedicle screwfixation. J Clin Neurosci 17:886-892, 2010.

9. Lozano AM, Vanderlinden G, Bachoo R, Rothbart P:Microsurgical C-2 ganglionectomy for chronic in-tractable occipital pain. J Neurosurg 89:359-365,1998.

0. Lu J, Ebraheim NA: Anatomic considerations of C2


1. Pan J, Li L, Qian L, Tan J, Sun G, Li X: C1 lateral massscrew insertion with protection of C1-C2 venous si-nus: technical note and review of the literature.Spine (Phila Pa 1976) 35:E1133-E1136, 2010.

2. Parker SL, McGirt MJ, Garces-Ambrossi GL, MehtaVA, Sciubba DM, Witham TF, Gokaslan ZL,Wolinksy JP: Translaminar versus pedicle screw fix-ation of C2: comparison of surgical morbidity andaccuracy of 313 consecutive screws. Neurosurgery64:343-348, 2009.

3. Payer M, Luzi M, Tessitore E: Posterior atlanto-axialfixation with polyaxial C1 lateral mass screws andC2 pars screws. Acta Neurochir (Wien) 151:223-229, 2009.

4. Poletti CE, Sweet WH: Entrapment of the C2 rootand ganglion by the atlanto-epistrophic ligament:clinical syndrome and surgical anatomy. Neurosur-gery 27:288-291, 1990.

5. Rocha R, Safavi-Abbasi S, Reis C, Theodore N, Bam-bakidis N, de Oliveira E, Sonntag VK, Crawford NR:Working area, safety zones, and angles of approachfor posterior C-1 lateral mass screw placement: aquantitative anatomical and morphometric evalua-tion. J Neurosurg Spine 6:247-254, 2007.

6. Saiki M, Taguchi T, Kaneko K, Toyota K, Kato Y, LiZ, Kawai S: Measuring of the compensation of anerve root in a cervical schwannoma: a case report. JOrthop Sci 8:714-716, 2003.

7. Squires J, Molinari RW: C1 lateral mass screw place-ment with intentional sacrifice of the C2 ganglion:functional outcomes and morbidity in elderly pa-tients. Eur Spine J 19:1318-1324, 2010.

8. Stechison MT, Mullin BB: Surgical treatment ofgreater occipital neuralgia: an appraisal of strate-gies. Acta Neurochir (Wien) 131:236-240, 1994.

9. Stulik J, Vyskocil T, Sebesta P, Kryl J: Atlantoaxialfixation using the polyaxial screw-rod system. EurSpine J 16:479-484, 2007.

0. Thomas JA, Tredway T, Fessler RG, Sandhu FA: Analternate method for placement of C-1 screws. JNeurosurg Spine 12:337-341, 2010.

1. Wang MY, Levi AD: Ganglionectomy of C-2 for thetreatment of medically refractory occipital neural-gia. Neurosurg Focus 12:E14, 2002.

onflict of interest statement: The authors declare that therticle content was composed in the absence of anyommercial or financial relationships that could beonstrued as a potential conflict of interest.

eceived 07 November 2010; accepted 08 July 2011;ublished online 07 November 2011

itation: World Neurosurg. (2012) 78, 1/2:170-177.OI: 10.1016/j.wneu.2011.07.010

ournal homepage: www.WORLDNEUROSURGERY.org

vailable online: www.sciencedirect.com

tion using plate and screw method: a report of 160treated patients. Neurosurgery 51:1351-1356, 2002.

nerve root ganglion. Spine (Phila Pa 1976) 23:649-652, 1998.

1878-8750/$ - see front matter © 2012 Elsevier Inc.All rights reserved.


http://www.worldneurosurgery.org

http://www.sciencedirect.com

world neurosurg 2011_ p170.pdf

Documents

uences of c2 nerve transection

nerve root sectioning

andellent clinical outcomes

elderly patients

patientsee patients

patients beforeticipating

instrumented fut

anaing outcomes