Third Nerve Palsy: Analysis of 1400

ORIGINAL ARTICLE

Personally-examined Inpatients
James R. Keane

ABSTRACT: Background: Most studies of third nerve palsy (TNP) antedate computerized imaging and focus primarily on chart
review of referral outpatients. Methods: To compare a large contrasting population, I reviewed 1400 personally-examined municipal
hospital inpatients with TNPs seen over 37 years. Results: TNPs were bilateral in 11%, complete in 33%, without other neurological
signs (isolated) in 36%, and associated with recurrent cranial neuropathies in 7%. Third nerve damage occurred in the subarachnoid
space in 32%, the cavernous sinus in 23%, the brainstem in 14%, as a nonlocalized peripheral neuropathy in 18% and at an uncertain
location in 13%. Causes were trauma (26%), tumor (12%), diabetes (11%), aneurysm (10%), surgery (10%), stroke (8%), infection
(5%), Guillain-Barre and Fisher syndromes (5%), idiopathic cavernous sinusitis (3%), benign self-limited (2%), miscellaneous (4%),
and unknown (3%). Local causes, besides an abundance of trauma, included six cases involving cysticercosis, four with wound
botulism, and one with coccidiomycotic meningitis. Of 234 patients with diabetes, microvascular ischemia was the cause of TNP in only
two-thirds (five had aneurysms) and 53% of those with diabetic microvascular ischemia had pupillary involvement—often bilateral,
suggesting concomitant autonomic neuropathy. Only 2% of aneurysms spared the pupil. A painful onset occurred with 94% of aneurysm
and 69% of diabetic cases. Conclusions: Bilateral TNPs, multiple cranial neuropathies, and accompanying neurological signs were
common among our inpatients, as were causes rare in outpatient settings such as severe trauma, transtentorial herniation, midbrain
strokes, and the Guillain-Barre syndrome. Few cases remained undiagnosed and nondiabetic ischemia was rare.

RÉSUMÉ: Paralysie du troisième nerf : analyse de 1400 patients hospitalisés examinés par le même observateur. Contexte : La plupart des études
sur la paralysie du troisième nerf (PTN) sont antérieures à la tomodensitométrie et sont constituées principalement de revues de dossiers de patients
externes. Méthodes : Afin d’examiner un volet différent de la PTN, j’ai révisé 1400 dossiers de patients atteints d’une PTN, hospitalisés dans un hôpital
municipal, que j’ai moi-même examinés au cours d’une période de 37 ans. Résultats : La PTN était bilatérale chez 11% des patients, complète chez
33%, sans autre signe neurologique (isolée) chez 36% et associée à des neuropathies crâniennes récurrentes chez 7%. La lésion du troisième nerf était
située dans l’espace sous-arachnoïdien chez 32%, dans le sinus caverneux chez 23%, dans le tronc cérébral chez 14%, comme une neuropathie
périphérique non localisée chez 18% et à un endroit indéterminé chez 13%. Les causes étaient un traumatisme chez 26%, une tumeur chez 12%, un
diabète chez 11%, un anévrisme chez 10%, une chirurgie chez 10%, un accident vasculaire cérébral chez 8%, une infection chez 5%, un syndrome de
Guillain-Barré ou de Fisher chez 5%, une sinusite caverneuse idiopathique chez 3%, une cause autolimitée bénigne chez 2%, des causes variées chez
4% et des causes inconnues chez 3%. Les causes locales, excluant les nombreux cas de traumatisme, comprenaient 6 cas de cysticercose, 4 de botulisme
relié à une plaie et 1 cas de méningite à coccidiomycose. L’ischémie microvasculaire était la cause de la PTN chez seulement les deux tiers des 234
patients diabétiques dont 5 étaient porteurs d’anévrismes et 53% avaient une atteinte pupillaire, souvent bilatérale, ce qui est compatible avec la présence
d’une neuropathie autonomique. La pupille n’était pas atteinte chez seulement 2% des patients porteurs d’anévrismes. Le début avait été douloureux
chez 94% des patients porteurs d’anévrismes et 69% des diabétiques. Conclusions : La PTN bilatérale, les neuropathies crâniennes multiples et les
signes neurologiques associés sont fréquents chez nos patients hospitalisés, de même que les étiologies qui sont rares chez les patients externes comme
les traumatismes sévères, la hernie transtentorielle, les accidents vasculaires cérébraux et le syndrome de Guillain-Barré. Un diagnostic n’a pu être posé
dans un très petit nombre de cas et l’ischémie non diabétique était rare.

Can. J. Neurol. Sci. 2010; 37: 662-670

Third nerve palsy (TNP) may be difficult to diagnose, the METHODS

causes are often elusive, the prognosis is potentially catastrophic, Inpatients personally-examined on wards of the Los Angeles
and the optimal workup is currently in flux. The landmark large County/University of Southern California Medical Center and
study of TNP is composed of four sequential articles on ocular
motor nerve palsies, each describing ten years experience at the
Mayo Clinic1-4. That survey and most other studies of TNP5-11
(Table 1) are based largely upon referral patients seen by many
different doctors with only several having even a token
experience with magnetic resonance imaging (MRI). The present From the University of Southern California Medical School, Los Angeles, California,
study explores the causes and sites of TNP in a large contrasting USA.

group of municipal hospital inpatients examined by a single

RECEIVED JULY 9, 2009. FINAL REVISIONS SUBMITTED MARCH 4, 2010.
Correspondence to: James R. Keane, 6281 Vine Way, Los Angeles, California, 90068,
neuro-ophthalmologist. USA.

662
Downloaded from https://www.cambridge.org/core. IP address: 36.72.212.135, on 25 Jan 2022 at 22:26:16, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/S0317167100010866
 LE JOURNAL CANADIEN DES SCIENCES NEUROLOGIQUES

Table 1: Causes and locations of third nerve palsy: present study

CAUSE number of cases (%) LOCATION number of cases

Cause Total Isolated Unilateral Bilateral Cavernous Subarachnoid Brainstem Peripheral Unknown
N=1400 N=500 N=1250 N=150 sinus Space N=197 nerve N=178
N=320 N=449 N=256
TRAUMA 369 (26) 107 (21) 354 15 97 112 27 0 133
TUMOR 173 (12) 11 (2) 154 19 112 42 18 0 1
DIABETES 156 (11) 152 4 0 0 0 156 0
151 (30)
ANEURYSM 145 (10) 101 (20) 144 1 8 134 3 0 0
SURGERY 140 (10) 40 (8) 134 6 38 79 23 0 0
STROKE 114 (8) 2 (0.4) 87 27 0 15 99 0 0
INFECTION 65 (5) 9 (2) 46 19 10 32 15 3 5
G-B syndrome 38 (3) 0 (0) 8 30 0 0 0 38 0
FISHER 24 (2) 0 (0) 2 22 0 0 0 24 0
syndrome
ICS 49 (3) 10 (2) 49 0 49 0 0 0 0
BENIGN 27 (2) 22 (4) 27 0 0 0 0 27 0

OTHER 63 (4) 17 (3) 56 7 4 35 11 8 5

(HERNIATION)Δ
UNKNOWN 37 (3) 30 (6) 37 0 2 0 1 0 34

(RECURRENT) Δ
[186 ] [19] [176] [9] [0] [159] [27] [0] [0]

ICS=idiopathic cavernous sinusitis, isolated=without other neurological signs; ΔCases of transtentorial herniation and recurrent
[92] [65] [87] [6] [22] [7 ] [3] [59] [2]

cranial neuropathies are tallied under primary causes

Rancho Los Amigos National Rehabilitation Center of Downey, 31, and videotapes of 54 were available), as well as selected
California, from 1971 through 2007 were studied. The Medical hospital charts and postmortem reports of 52 patients. Diagnosis
Center, a large municipal hospital serving an urban population and location were determined by history and examination,
with many recent immigrants, shares patients with other local supported by laboratory studies, testing for myasthenia,
municipal hospitals and clinics, ethnic healers and pharmacies, cerebrospinal fluid examination, and generous use of
and medical facilities in Mexico. (Each year nearly one million contemporary radiological contrast studies and computerized
residents of Los Angeles seek medical care in Mexico12.) As scans, as clinically indicated. About 100 patients antedated
advanced disease is common and follow-up limited, most computed tomogram (CT) scanning and 650 preceded use of
patients with TNP are evaluated as inpatients. MRI at our hospital.
The present study is based upon my notes of all patients and Involvement of two or more elements of the third nerve (TN)
photographs of many (slides of 555 patients, super 8 film clips of was required for inclusion and maximum observed severity was
used for comparison. Only the immediate neuropathy of serial
TNPs was tallied. Comatose patients lacking spontaneous eye
movements were excluded. When multiple causes or locations
applied, the single most important one was selected. A “benign”
Table 2: Cranial neuropathies accompanying third nerve diagnosis was applied to self-limited TNPs in patients under age
palsy 50 if extensive testing revealed no other cause. Botulism was
included but myasthenia was not.
Cranial nerve number #Nerves (%) N=1095
2 158 (14) RESULTS
4 138 (13) Of 1400 patients, 792 (57%) were male and ages ranged from
3 to 89 years with a mean of 42. The right side was involved in
5 155 (14)
654 (47%), the left in 596 (43%), and both sides in 150 patients
6 336 (31) (11%), giving a total of 1550 impaired third nerves (TNs). The
7 184 (17) subarachnoid space was the site of involvement in 32% of cases,
8 27 (2) the cavernous sinus in 23%, the brainstem in 14%, a non-
localized neuropathy in 18%, and location was uncertain in 13%
10 53 (5) (Table 1).
11 15 (1) Third nerve palsy was the sole neurological sign (isolated) in
12 29 (3) 500 cases (40% of unilateral cases and 4% of bilateral cases)

Volume 37, No. 5 – September 2010 663

Downloaded from https://www.cambridge.org/core. IP address: 36.72.212.135, on 25 Jan 2022 at 22:26:16, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/S0317167100010866
 THE CANADIAN JOURNAL OF NEUROLOGICAL SCIENCES

Table 3: Prior cranial neuropathies in patients with third nerve palsy

Cranial Number Of Cranial Neuropathies

Nerve

Diabetes Benign ICS Infection Tumor Aneurysm Vasculitis G-B-F Ophthalmic Other Total
N=41 N=16 N=16 N=2 N=5 N=4 N=2 N=2 Migraine ** N=93
N=2 N=3

2 0 0 0 1 1 2
3 14 4 4 2 2 10 1 37
4 1 0 0 1
5 0 1 1 1 3
6 11 10 10 3 1 1 0 1 37
7 30 2 2 1 2 2 3 42
8 0 0 0 1 1 1 3
10 0 0 0 0
11 0 0 0 0
12 0 0 0 0

N=Number of patients, *cysticercosis, syphilis; **trauma, benign intracranial hypertension, unknown

TOTAL 56 17 17 3 6 4 4 2 10 6 125

(Table 1) Isolated cases occurred most commonly with diabetic The pupil was fixed in 40%, either fixed or minimally
TNP (97%), benign cases (81%), unknown causes (81%), and responsive (fixed to the casual examiner) in 54%, moderately
aneurysms (70%), and least commonly with the Guillain-Barre reactive in 17%, normally reactive in 24% (normally reactive
(G-B) and Fisher (F) syndromes (0), strokes (2%), other (3%), and equal [spared] in 14%), and reactivity was obscured by
and tumor (6%) (Table 1). The TN was the only cranial nerve iatrogenic dilatation or surgery in 5% (Table 5).
affected in 843 patients (60%). In the 557 patients with 1095 The pupil on the side of the TNP was ≥0.5mm smaller in 45
concomitant neuropathies of other cranial nerves, sixth nerve cases (3% of unilateral cases). None of these patients showed
palsies were common; seventh, second, fifth, and fourth evidence of TN aberrant regeneration, but six had Horners
neuropathies occurred frequently; and tenth, twelfth, eighth, and syndrome. The light reaction was impaired in 23 patients
eleventh nerve palsies were rarely seen (the first and ninth nerves (bilaterally in nine). Miosis occurred with all locations and most
were not tallied) (Table 2). causes but was singularly common (15%) in idiopathic
Recurrent cranial neuropathy was a feature of 93 patients with cavernous sinusitis cases (Table 5).
125 prior neuropathies. Nearly one-half of recurrences involved An oval pupil, found in 49 cases (55 eyes), also was present
diabetic TNPs (41 cases, 56 nerves), and a majority of the in all locations and most causes but was singularly common
remaining cases were divided between idiopathic cavernous (17%) in Fisher’s syndrome (Table 5). There was no overlap of
sinusitis (ICS) and benign causes (Table 3). Most patients (66) miotic and oval pupils.
had two episodes of resolving cranial neuropathy, but 21 had The pupil in 117 TNPs caused by pcom aneurysms was
three, six had four, and one patient with ophthalmoplegic reactive in 37% but normal in just 1% (the case with superior
migraine syndrome had six episodes. In ten cases (five tumor, division palsy excepted) (Table 5). The pupil was abnormal in
three aneurysm, and one each G-B and F syndromes) prior
neuropathies were unrelated to the current TNP and most often
were due to Bell’s palsy (four cases) or old head trauma (two
cases).
Despite limited follow-up, signs of aberrant regeneration of Table 4: Causes of third nerve palsy with aberrent
the TN were noted in 95 patients. Most cases were due to trauma regeneration
(65%), with aneurysms (13%), tumors (8%) and surgery (8%)
comprising most of the remaining cases (Table 4). A superior
division TNP was seen in 20 cases, with three each due to
Cause # patients % with % of aberrant

surgery, diabetes and trauma (two herniation), two each

N=95 same regeneration

aneurysm and benign causes, and other causes seven. A single

diagnosis cases

patient had an inferior division palsy, following trauma.

TRAUMA 62 17 65

Associated neuro-ophthalmological findings included

ANEURYSM 12 8 13

nystagmus in 7%, homonymous hemianopia in 7%, gaze pareses

TUMOR 8 5 8

in 4%, papilledema in 3%, and internuclear ophthalmoplegia

SURGERY 8 6 8

in 1%.
OTHER 5 1 5
CAUSES

664
Downloaded from https://www.cambridge.org/core. IP address: 36.72.212.135, on 25 Jan 2022 at 22:26:16, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/S0317167100010866
 LE JOURNAL CANADIEN DES SCIENCES NEUROLOGIQUES

Table 5: Pupil findings vs. cause and location of third nerve palsy
Patients (%)
ETIOLOGY FIXED FIXED & LIMITED SPARED SMALLER OVAL
PUPIL MINIMALLY REACTION
REACTIVE OR UNEQUAL
TRAUMA N=364 * 59 72 19 9 2 3
TUMOR N=158 39 57 34 9 4 5
DIABETES 10 27 26 47 7 1
N=141
ANEURYSM 61 77 21 2 3 2
N=143
SURGERY N=124 46 59 35 6 2 2
STROKE N=110 43 59 27 14 2 5
INFECTION 48 68 23 9 3 8
N=65
G-B SYNDROME 24 38 27 35 0 6
N=34
FISHER 38 50 25 25 0 17
SYNDROME
N=24
ICS N=46 13 26 52 22 15 0
BENIGN N=25 0 8 44 48 4 0
OTHER N=60 40 58 29 13 3 10
UNKNOWN N=34 3 15 38 47 3 0
TOTAL N=1333 42 57 28 15 3 4
[HERNIATION 42 59 32 9 4 4
N=184
[RECURRENT 12 25 42 33 5 0
N=88
LOCATION
CAVERNOUS 46 59 30 11 2 5
SINUS N=295
SUBARACHNOID 46 63 30 7 3 3
SPACE N=436
BRAINSTEM 48 66 22 12 3 5
N=190
NERVES N=164 13 30 34 36 5 2
UNKNOWN 43 56 23 21 2 3

*N=number of cases with pupils unobscured by prior surgery or iatogenic dilatation

N= 248

53% of diabetic TNPs and in 59% of these, impairment was Trauma, the cause of 26% of all TNPs, was the most common
bilateral. Although only a few pupils were pharmacologically etiology by far, with vehicle-related injuries accounting for one-
tested--with positive results, others displayed light-near half of the cases (36% automobile, 11% pedestrian, and 3%
dissociation, and autonomic neuropathy was a likely bicycle) (Table 6). Despite presenting a small target, the TN was
concomitant in many patients. A fixed pupil opposite a normal directly injured by penetrating trauma in 33 cases including one
pupil was rare with diabetic TNPs (3%), common with TNPs due remarkable patient with an MRI-documented track of a
to pcom aneurysms (61%), and intermediate with TNPs caused sharpened screwdriver driven into the skull, through the
by tumors (29%). Complete pupillary sparing was most common cerebrum, to damage the TN fascicles in the cerebral peduncle14
with benign cases (48%), diabetes (47%), unknown causes (Table 6). Trauma was documented in 69% of cases of TNP
(47%) and the G-B (35%) and F (25%) syndromes, and least caused by transtentorial herniation and subdural hematoma was
common with aneurysms (2%), surgery (6%), tumor (9%), the proximate cause of 67% of all tentorial herniations.
infection (9%) and tentorial herniation (9%) (Table 5). Tumors causing TNP were more often malignant (100 cases)
A unique pupillary phenomenon occurred in a 26-year-old than benign (71 cases) and nearly two-thirds involved the TN in
man with a traumatic complete TNP in whom the affected, light- the cavernous sinus (Tables 1, 7). Common tumors included
fixed pupil rhythmically expanded and contracted once a second, pituitary adenomas (24%, more than half apoplectic), solid
unaccompanied by movement of the globe or the opposite tumor metastases (21%), and lymphomas (16%--more than one-
normal pupil13. third associated with acquired immune deficiency syndrome)
Among unilateral cases, ptosis was complete in 54%, severe (Tables 1, 7). Surgery accounted for 10% of TNPs, with 87
in 8%, moderate in 17%, mild in 17%, normal in 2%, and aneurysm cases (30 basilar) easily outnumbering the 44 tumor
unknown in 2%. Overall, elevation (absent in 61% of unilateral surgeries (headed by 21 meningiomas and 12 pituitary
cases and 66% of all TNPs) was slightly more affected than adenomas) (Table 8).
adduction (absent in 55% and 59%) or depression (absent in 52% Diabetes was present in 234 cases (17% of all patients), but
and 57%), but the differences were modest and of no differential diabetic nerve ischemia was the cause of TNP in only 156 (two-
value.

Volume 37, No. 5 – September 2010 665

Downloaded from https://www.cambridge.org/core. IP address: 36.72.212.135, on 25 Jan 2022 at 22:26:16, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/S0317167100010866
 THE CANADIAN JOURNAL OF NEUROLOGICAL SCIENCES

Table 6: Sources of trauma causing third nerve palsy Table 7: Tumors causing third nerve palsy

Total Direct Secondary Tumor Type # Patients

# cases trauma Herniation
PRESENT STUDY MAYO1-4 WILLS5
AUTOMOBILE 134 111 23 PITUITARY ADENOMA 41 (22 apoplectic) 70
METASTASES (SOLID 36 108 4
FALL 60 20 40 TUMOR)
LYMPHOMA 23 (8 AIDS)
PEDESTRIAN 39 33 6 MENINGIOMA 16 60 1
NASOPHARYNGEAL 14 12
GUNSHOT WOUND 33 28 5
CARCINOMA
MOTORCYCLE 29 20 9 GLIOMA 12 7
LEUKEMIA 6
BEATING 30 10 20 SARCOMA 4
SCHWANNOMA* 4
BICYCLE 10 5 5 CRANIOPHARYNGIOMA 4
CHORDOMA/ 4 10
STABBING 5 3 2 CHONDROMA
#
OTHER 9 43
HORSE 5 4 1

*2 of 8th nerve and 2 of 5th nerve origin; #2 plasmacytoma, 2

TOTAL 173 310 5
OTHER* 4 3 1
unknown. 1 each germinoma, primitive neuroectodermal tumor, oligo-
dendroglioma, pinealoma, epidermoid
UNKNOWN 20 5 15
TOTAL 369 242 127

*Soccer heading, car falling from jack, moped and tractor accidents

thirds of all diabetics and 11% of all TNPs). In the remaining 78

diabetics, the etiology of TNP was potentially related to diabetes
in about half (stroke 24, infection 16). In five diabetics, pcom Table 8: Operations causing third nerve palsy
aneurysms were the cause of TNP. Pain was a presenting
symptom in 69% and eye movements were completely affected
in 51%. Pupillary abnormalities (53%, 31% bilateral) were ANEURYSM 87
common but anisocoria greater than 2mm was seen in only 3%.
POST. COMMUNICATING ARTERY 36
Despite extensive use of MRI, only a single case of an apparent
(partial) diabetic TNP proved to be due to midbrain infarction15. BASILAR ARTERY 30
The 145 TNPs caused by aneurysms included 20% of isolated OTHER LOCATIONS 21
TNPs but just 5% of cases with other neurological signs. Only
6% (four with cavernous and three with supraclinoid carotid TUMOR 44
aneurysms) of the 120 patients able to give a history had no pain. MENINGIOMA 21
The supraclinoid carotid artery was the site of 124 aneurysms,
PITUITARY ADENOMA 12
with 117 being true pcom aneurysms (Table 9). Middle-aged
patients harbored most aneurysms, but three patients under the CHORDOMA 4
age 20 (the youngest, with coarctation, was 11) and two over 75 CRANIOPHARYNGIOMA 3
had pcom aneurysms. Of 78 patients with pcom aneurysms who
had complete involvement of eye movements, the pupil was SCHWANNOMA 2
fixed in 66, minimally reactive in nine, and had limited reaction EPIDERMOID 2
in the remaining three. The pupils were spared in only a single
ARTERIOVENOUS MALFORMATION 3
patient (with partial eye movement limitation).
Six of nine cavernous carotid aneurysms had additional CAROTID-CAVERNOUS FISTULA 2
cranial neuropathies. The seven TNPs caused by distal basilar OTHERS* 4
aneurysms were overshadowed by 30 TNPs which followed
basilar aneurysm surgery. In two cases, the onset of TNP TOTAL 140
coincided with subarachnoid hemorrhage from an aneurysm
remote from the TN (anterior cerebral and a peripheral convexity *1 each cysticercosis, ethmoiditis, abscess drain insertion, cavernous
mycotic aneurysm). No cause of bleeding could be determined in angioma
seven cases of subarachnoid hemorrhage.

666
Downloaded from https://www.cambridge.org/core. IP address: 36.72.212.135, on 25 Jan 2022 at 22:26:16, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/S0317167100010866
 LE JOURNAL CANADIEN DES SCIENCES NEUROLOGIQUES

Table 9: Description and pupillary signs of TNPs due to aneurysms (number of cases)

Description Pupil reactions

LOCATION ALL GIANT MYCOTIC FIXED MINIMAL LIMITED NORMAL/ NORMAL OBSCURED*
NORMAL
DILATED
POST 117 2 0 76 13 19 6 1 + 1** 1 1
COMMUNICATING
A.
CAVERNOUS 9 2 0 5 2 2 0 0 0
CAROTID A.
OTHER CAROTID A. 7 4 0 3 2 1 0 0 + 1** 1
BASILAR A. 7 1 1 5 1 0 1 0 0
POSTERIOR 1 0 1 1 0 0 0 0
CEREBRAL A.
ANTERIOR 1 0 0 0 0 0 0 1 1
CEREBRAL A.
PERIPHERAL A. 1 0 1 1 0 0 0 0 0
UNKNOWN 2 0 0 1 0 0 0 0 1 0
TOTAL 145 9 3 92 18 22 7 2 + 2** 2 3

*iatrogenically-dilated pupils; ** superior division palsies

Primary midbrain strokes were responsible for 7% of all TNPs Tentorial herniation without known trauma resulted in 47
and 18% of bilateral TNPs. Of the 96 cases, 82% were infarcts, TNPs. Overall, herniation caused 186 TNPs in association with
five associated with vasculitis (Table 10). Midbrain hemorrhage 95 subdural hematomas, 44 epidural hematomas, 33
in 19 cases (13 associated with vascular malformation), parenchymal hemorrhages, 5 tumors, 4 infarcts, 2 cerebral
originated in the midbrain in ten, dissected from the thalamus in abscesses, 1 epidural abscess, and 2 unknown. Bilateral subdural
six and from the pons in three. In only two midbrain strokes were hematomas were the result rather than the cause of herniation in
TNPs the sole neurological sign16. a confusing patient with headache and a TNP which switched
sides after several days. Magnetic resonance imaging showed
signs of spontaneous intracranial hypotension and he promptly
responded to an intrathecal blood patch.
In 65 cases of infection, the meninges were principally
involved in nearly half (48%), the brainstem in 23%, and
Table 10: Vascular causes of third nerve palsy cavernous sinus in 17% (Table 11). Nearly one-fourth of all cases
were due to bacterial meningitis, but a wide range of infections
were represented. Acquired immune deficiency syndrome was
present in 19 patients, of whom 8 had lymphoma (meningeal in
6), 6 had toxoplasmosis, and 5 had cryptococcal meningitis.
CAUSE # CASES

Cysticercosis is the most common nervous system infection on

ANEURYSM 145

our neurology service but was associated with TNPs in only six
cases: Three resulted from midbrain cysts, one was due to
MIDBRAIN STROKE 96

cysticercal meningitis, and two were indirect: one a secondary

Infarct* 77

vasculitic infarction and the other resulted from surgical removal

Hemorrhage** 19

of a cysticercal cyst. Other diseases of local importance included

wound botulism (four cases) and coccidiomycosis (one case).
VASCULAR CAUSES OF HERNIATION 47

Syphilis and tuberculosis continue to be occasional causes of

(no history of trauma)

TNP in our municipal hospital.

Subdural Hematoma 29

Potentially dangerous delay in diagnosis occurred with two

Cerebral Hemorrhage 14

cases of meningitis. Both presented with partial TNPs several

Cerebral Infarct 4

weeks after the onset of headache. After lumbar puncture showed

cerebrospinal fluid pleocytosis with a few red cells but no
MISCELLANEOUS VASCULAR CAUSES 16

xanthochromia, diagnoses of subacute infectious meningitis

Subarachnoid Hemorrhage, Unknown Cause 7

were made. Progression of TNPs led to the discovery of

Carotid-Cavernous Fistula *** 2

causative pcom aneurysms. In retrospect, cerebrospinal fluid

changes un-doubtedly reflected a late meningeal reaction to
Microvascular Ischemia 7

*Includes 5 cases of vasculitis; **Includes 10 arteriovenous malformations, subarachnoid hemorrhage17.

TOTAL 304

3 cavernous angioma, 1 pregnancy induced hypertension, 1 idiopathic The G-B and Fisher syndromes (62 patients) occasionally
thrombocytopenic purpura; ***additional cases tallied under surgery (2) present with isolated TNPs, but a more diffuse ophthalmoparesis
and trauma (1) is usual: 84% had bilateral TNPs and 32% had complete bilateral

Volume 37, No. 5 – September 2010 667

Downloaded from https://www.cambridge.org/core. IP address: 36.72.212.135, on 25 Jan 2022 at 22:26:16, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/S0317167100010866
 THE CANADIAN JOURNAL OF NEUROLOGICAL SCIENCES

Table 11: Infectious causes of third nerve palsy (number of cases)

CAUSE LOCATION
TOTAL MENINGITIS BRAINSTEM CAVERNOUS UNKNOWN/
SINUS DIFFUSE
BACTERIA 19 15 (2 solitary) 2 (solitary) 2
MUCORMYCOSIS 7 7
BOTULISM (wound 4) 7 7
TOXOPLASMOSIS 6 6 (6 AIDS)
CRYPTOCOCCUS 6 6 (5 AIDS: 3 solitary)
TUBERCULOSIS 4 2 1 1
SYPHILIS 4 2 2 (1 solitary)
CYSTICERCOSIS 4 1 3
HERPES ZOSTER 2 1 1
COCCIDIOMYCOSIS 1 1
ASPERGILLIS 1 1
CANDIDA 1 1 (solitary)
UNKNOWN 3 2 1
TOTAL 65 31 15 11 8

ocular paralysis. By contrast, idiopathic cavernous sinusitis syndrome, and an undiagnosed neuro-degeneration with
(Tolosa-Hunt syndrome) (ICS) (49 cases), presented with an parkinsonian features. In 37 cases the cause was unknown.
isolated TNP in 20% and complete unilateral ocular paralysis in
just 4%. Indeed, damage to the sixth (12 cases) and fourth (7 DISCUSSION
cases) nerves was much less common with ICS than impairment History
of the second (37 cases) and fifth (26 cases) nerves. Benign self-
limited cases, often recurrent, compose 2% of this series. Whereas Galen (130-201) believed that the oculomotor nerve
Other causes/associations (63 cases) include subdural (the 2nd of his seven cranial nerves) supplied all of the
hematoma causing tentorial herniation without evidence of extraocular muscles, Willis (1621-1675) established the
trauma 29, subarachnoid hemorrhage of unknown cause 7, oculomotor as the third cranial nerve (of his eight) and connected
nondiabetic microvascular ischemia 7, congenital 5, spontaneous the 3rd, 4th, and 6th nerves with the appropriate muscles18.
carotid cavernous fistula 2, ophthalmoplegic migraine syndrome In 1888, Gowers was aware that tumors, meningitis and
2, acute hydrocephalus of unknown origin 2, and one each aneurysms caused TNP and considered “rheumatic neuritis” due
spontaneous intracranial hypotension, lymphomatoid granulo- to cold drafts as an unproven but “highly probable”cause19. Less
matosis, histiocytosis, acute demyelination, Leigh’s disease, than 20 years later, a dozen cases of diabetic ocular palsy had
Chronic Inflammatory Demyelinating Polyneuropathy, severe been recognized, myasthenia had emerged from the catch-all
basilar artery ectasia, benign intracranial hypertension diagnosis of “nuclear ophthalmoplegia,” and belief in rheumatic

Table 12: Comparison of third nerve palsy series

PRESENT STUDY MAYO CLINIC1-4 BERLIT11 GREEN,6 LAERE8 Combined

Smaller
Series
5,7,9,10

CAUSE TOTAL ISOLATED TOTAL ISOLATED ISOLATED TOTAL TOTAL TOTAL

N=1136
N=1400 N=500 N=1491 N=172 N=130 N=109 N=236
TRAUMA 26% 21% 15% 15 6% 11% 11% 8%
TUMOR 12 6 14 16 7 4 4 19
DIABETES 11 26 --* --* 42 19 1 25
ISCHEMIC <1 <2 13* 17* 7 15 --- 9
(NONDIABETIC)
ANEURYSM 10 21 3 3 9 29 17 16
SURGERY 10 8 --- --- 0 2 0
STROKES + 8 1 17 12 ----- 33 3
INFECTION 5 2 --- --- 8 9 --- ---
G-B syndrome 3 0 --- --- 0 --- ---
FISHER synd 2 0 --- --- 0 --- ---
ICS 3 2 --- --- 0 --- ---
BENIGN 2 4 --- --- -- 6
OTHER 4 2 16 14 11 4 6 14

*diabetic and nondiabetic microvascular ischemic cases grouped together

UNKNOWN 3 5 21 24 10 8 28 9

668
Downloaded from https://www.cambridge.org/core. IP address: 36.72.212.135, on 25 Jan 2022 at 22:26:16, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/S0317167100010866
 LE JOURNAL CANADIEN DES SCIENCES NEUROLOGIQUES

palsies persisted20. In the century since, insulin, antibiotics, and it’s likely that our inpatient diabetics are sicker than the usual
cerebral angiography, and the automobile have affected the rank ischemic TNP patient. Our 53% incidence of pupillary
order of TNP causes and rheumatic palsy slowly became extinct. involvement with diabetic TNP is considerably higher than the
usual 10-40% range and more than half of our cases with
Analysis of cases affected pupils had bilateral involvement, probably reflecting
The present study reflects a municipal hospital’s heavy associated diabetic autonomic neuropathy. An additional
burden of poverty, trauma, and neglected disease. A few cases of possibility, perhaps unique to a municipal hospital population, is
cysticercosis are the principal reminder of the large immigrant the high frequency of repeated--and often forgotten—eye
population. Bilateral TNPs21, multiple cranial neuropathies22, trauma, which could provide a cause for an occasional
and accompanying neurological signs are common among our unexplained pupillary abnormality29.
inpatients, as are causes seldom seen in outpatient settings such Idiopathic cavernous sinusitis is mentioned only occasionally
as severe trauma, midbrain strokes, and the G-B syndrome. in TNP series. The cavernous sinus was relatively blind to
Divisional palsies of the TN have long been recognized23, and radiological investigation before modern imaging and it’s likely
two cases caused by aneurysms were included in an early TNP that some cases of “sinusitis” or “presumed parasellar
series6. Once thought to strictly parallel TN anatomy and localize aneurysms”5 actually had ICS. Although it remains a diagnosis
in or near the orbit, division palsies have repeatedly been of exclusion which can be confused with lymphoma and other
demonstrated with lesions in the brainstem and subarachnoid tumors, and atypical presentations led to surgical exploration in
space. Myasthenia often mimics superior division palsies as, on two of our cases, MRI diagnosis of ICS has proven surprisingly
occasion, do lesions of the orbital roof24. accurate30.
Recurrent cranial neuropathies usually signify benign disease, While antibiotics have demoted infection from the principal
but 18 of the 93 patients in this study with recurrent etiologies of TNPs, it continues to be a frequently obscure and
neuropathies—4 with aneurysms-- had ominous conditions. One singularly treatable cause. Among our patients, cysticercosis
in seven patients had a prior neuropathy unrelated to the present reflected immigration from Mexico and Central America,
condition; in particular, previous seventh nerve palsy does not coccidiomycosis resulted from residency in the Southwestern
guarantee benignity. deserts, wound botulism was caused by “skin-popping” drugs
Smaller pupils--in 3% of our patients--accompany TNP under unsterile conditions, and syphilis and tuberculosis
occasionally in other series5. Horners syndrome and diabetic reflected our patients’ culture of poverty.
autonomic neuropathy could explain about one-quarter of our Multiple sclerosis (MS) was mistakenly considered a
cases. Physiological anisocoria (≥0.5mm), seen in 9% of normal common cause of TNP before internuclear ophthalmoplegia
pupils on single observations25, could account for a few more became recognized. While it’s an occasional cause of sixth nerve
cases, but at least nine cases had no ready explanation. Pupils in palsy, it’s curiously rare as a cause of TNP. Only one of our cases
third nerve palsy due to cavernous sinus lesions are said to be had MS (vs acute demyelinating encephalomyelitis) and there is
relatively smaller because of co-existing sympathetic deficits but a dichotomy in the reported incidence with many TNP studies
pharmacological testing has not supported that thesis26. In the having a single case of MS or none at all1,5,6,9,10 and others
present study, miosis was relatively uncommon with lesions in reporting a 2-3% incidence2,3,7,11.
the cavernous sinus overall, but occurred in 15 % of our patients
with ICS (Table 5). Other studies
The 1% incidence of pupillary sparing in our cases of pcom The hallmark Mayo Clinic studies of ocular motor palsies,
aneurysms is similar to the 2-3% in previous studies1,2,5 but encompassing a 40 year period through 1988, included 1491
lower than some studies utilizing neurosurgical chart review, a patients with TNP1-4 (Table 12). Most patients had isolated TNPs
process which may overlook some patients with subtle pupillary (76% vs 36% in the present study) and many remained
abnormalities27. undiagnosed (21% vs 3%). Trauma, although less common (15%
Among the more mysterious of TNPs are those—like two in vs 26%), was still a major cause. Aneurysms (14% vs 10%) and
our series--which coincide with aneurysm ruptures remote from tumors (17% vs 12%) were somewhat more common at Mayo
the TN. An early TNP study included a ruptured middle cerebral but infections (~3% vs 5%) less so.
artery aneurysm5, and a number of case reports of remote Other reports of TNP prior to the widespread use of computed
aneurysms at different sites have been published. The obvious scanning include experience at the Massachusetts Eye and Ear
explanation in such cases, as in our seven cases without a source Infirmary (1960)5 (59 patients:19% aneurysms and 34% diabetic
of subarachnoid hemorrhage, would be that the bleeding TNP), the Wills Eye Hospital neurology clinic study (1964)6
aneurysm was destroyed during rupture and the remote (130 patients: 29% aneurysms and 19% diabetes), a series from
aneurysm was incidental. However, evidence in our cases and Washington University (1965)7 (53 patients: ischemia 28%,
others, from imaging and surgery, is against that premise. Other aneurysm 19% and tumor 19%), and a study from a Belgium
explanations are highly speculative, but a few patients (like our neurosurgery clinic (1971)8 (109 patients: aneurysms 36%,
mycotic aneurysm case)--could possibly represent a “double hit” tumors 16%, and unknown 27%) (Table 11).
in which a TN weakened by meningitis develops a manifest Later studies utilizing CT scans included a report from a
palsy in the face of sudden severe increased intracranial university neurology department (1985)9 (61 patients with
pressure28. isolated TNPs: ischemia 38%, tumor 21%, trauma 15%), a
Despite our hospital’s low admission threshold, university neuro-ophthalmology series featuring a cliniconeuro-
microvascular ischemia is under-represented in the present study radiologic approach (1987)10 (63 patients: miscellaneous 19%,

Volume 37, No. 5 – September 2010 669

Downloaded from https://www.cambridge.org/core. IP address: 36.72.212.135, on 25 Jan 2022 at 22:26:16, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/S0317167100010866
 THE CANADIAN JOURNAL OF NEUROLOGICAL SCIENCES

ischemic 16%, tumor 16%, and undiagnosed 2%), and a review 20. Duane A. The extra-ocular muscles. In: Posey WM, Spiller WG,
of ocular motor palsy among university hospital inpatients editors. The eye and the nervous system. Philadelphia: JB
Lippincott Co; 1906. p. 222-41.
(1990)11 (218 patients altogether; of 172 with isolated TNPs: 21. Keane JR. Bilateral involvement of a single cranial nerve. Analysis
42% diabetes, 10% unknown, 9% aneurysm). of 578 cases. Neurology. 2005;65:950-2.
The present study is the first to utilize MRI frequently. Early 22. Keane JR. Multiple cranial nerve palsies: analysis of 979 cases.
MRI technique was not sufficiently sensitive to detect all Arch Neurol. 2005;62:1714-17.
aneurysms causing TNPs31. Now, however, the best modern 23. Williams HW. The diagnosis and treatment of diseases of the eye.
Boston: Houghton, Mifflin & Co; 1881. p. 310.
magnetic resonance angiography (MRA) and computarized 24. Keane JR. Ptosis and levator palsy caused by orbital roof fractures:
tomographic angiography equipment, in experienced hands, can three cases with subfrontal epidural hematomas. J Clin
detect all aneurysms of sufficient size to cause TNP, with Neuroophthalmol. 1993;13:225-8.
minimal complications and at a nearly-reasonable cost. In 25. Lam BL, Thompson HS, Corbett JJ. The prevalence of simple
anisocoria. Am J Ophthalmol. 187;104:69-73.
evaluating TNPs of uncertain cause, both diagnostic catheter 26. Trobe JD. Glaser JS, Post JD. Meningiomas and aneurysms of the
angiography and watchful waiting should soon give way to cavernous sinus: neuro-ophthalmologic features. Arch
prompt MRI/MRA screening. Ophthalmol. 1978;96:457-67.
27. Keane JR. Aneurysms and third nerve palsies. Ann Neurol. 1983;
ACKNOWLEDGEMENT 14;696-7.
28. Keane JR. Intermittent third nerve palsy with cryptococcal
The author thanks Frances C. Keane, RN, MPH, who helped meningitis. J Clin Neuroophthalmol. 1993;13:124-6.
with data retrieval and organization. 29. Keane JR. Functional diseases affecting the cranial nerves. In: J
Noseworthy, editor. Neurological therapeutics: principles and
practice. 2nd ed. London: Martin Dunitz; 2006. p. 2202.
REFERENCES 30. Keane JR. Cavernous sinus syndrome: analysis of 151 cases. Arch
1. Rucker CW. Paralysis of the third, fourth and sixth cranial nerves. Neurol. 1996;53:967-71.
Am J Ophthalmol. 1958;46:787-94. 31. Keane JR. Third nerve palsies and angiography. Arch Neurol. 1991;
2. Rucker CW. The causes of paralysis of the third, fourth and sixth 48:470.
cranial nerves. Am J Ophthalmol. 1966;61:1293-8.
3. Rush JA, Younge BR. Paralysis of cranial nerves III, IV, and VI:
cause and prognosis in 1000 cases. Arch Ophthalmol. 1981;99:
76-9.
4. Richards BW, Jones FR Jr, Younge BR. Causes and prognosis in
4278 cases of paralysis of the oculomotor, trochlear, and
abducens cranial nerves. Am J Ophthalmol. 1992;113:489-96.
5. Goldstein JE, Cogan DG. Diabetic ophthalmoplegia with special
reference to the pupil. Arch Ophthalmol. 1960;64:592-600.
6. Green WR, Hackett ER, Schlezinger NS. Neuro-ophthalmologic
evaluation of oculomotor nerve paralysis. Arch Ophthalmol.
1964;72:154-67.
7. Cheek CW, Simon KA, Gay AJ. Acquired cranial nerve lesions
affecting the ocular system. Am J Ophthalmol. 1965;59:13-7.
8. van Laere E. La paralysie non traumatique du nerf oculomoteur
commun. A propos de 109 observations. Acta Neurol Belg. 1974;
74:182-92.
9. Teuscher AU, Meienberg O. Ischaemic oculomotor nerve palsy;
clinical features and vascular risk factors in 23 patients. J Neurol.
1985;232:144-9.
10. Kwan ESK, Laucella M, Hedges TR jr, et al. A cliniconeuro-
radiologic approach to third cranial nerve palsies. AJNR Am J
Neuroradiol. 1987;8: 459-68.
11. Berlit P. Isolated and combined pareses of cranial nerves III, IV, and
VI. A retrospective study of 412 patients. J Neurol Sci. 1991;
103:10-5.
12. Wallace SP, Mendez-Luck C, Castafieda X. Heading south: why
Mexican immigrants in California seek health services in
Mexico. Med Care. 2009:47:662-9.
13. Keane JR. Rhythmic pupillary oscillations accompanying a
complete third nerve palsy. J Clin Neuro-ophthalmol. 1989;9:
169-70.
14. Keane JR. Third nerve palsy due to penetrating trauma. Neurology.
1993:43:1523-7.
15. Keane JR, Ahmadi J. Most diabetic third nerve palsies are
peripheral. Neurology. 1998;51;1510.
16. Keane JR. Isolated brain stem third nerve palsy. Arch Neurol. 1988;
45:813-4.
17. Keane JR. Aneurysmal third nerve palsy presenting with
pleocytosis. Neurology. 1996:46:1176.
18. Rucker CW. History of the numbering of the cranial nerves. Mayo
Clin Proc.1966;41:453-61.
19. Gowers W. Diseases of the nervous system. Philadelphia: JB
Lippincott Co; 1888. p. 612-16.

670
Downloaded from https://www.cambridge.org/core. IP address: 36.72.212.135, on 25 Jan 2022 at 22:26:16, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/S0317167100010866