Oculomotor nerve palsies are the most common if not the only cranial neuropathy observed in patients with diabetes.
Ogle in 1866, was the first author to mention the occurrence of diabetic ophthalmoplegia (3). In 1905, Dieulafoy published a series of 58 personal cases, in which most of the clinical characteristics of diabetic ophthalmoplegia were described (4). In 1935, Waite and Beetham (5) performed the first epidemiological study on the subject in which they compared the occurrence of oculomotor palsy in 2002 diabetic patients with 457 patients without diabetes. A series of other clinical reports have refined our knowledge on the subject but pathological studies remain scanty with only a few autopsy cases studied (6-8) and the pathophysiology of oculomotor palsies in patients with diabetes remains a matter of discussion.
Such as focal neuropathy observed in other sites of the body, diabetic ophthalmoplegia is uncommon in diabetic patients. In 1933, Gray (9) observed two patients with ophthalmoplegia among 500 diabetic patients examined and Waite and Beetham (5,10) estimated the incidence of oculomotor palsy among patients with diabetes to be 0.8-1.8%. It is interesting to note that in this study, the frequency of oculomotor palsy was 0.8% in patients of less than 45 years of age, against 2.1% after 45 years.
Frequency of involvement of the different oculomotor nerves: the sixth and the third cranial nerves are most commonly affected. In a series of 58 cases of diabetic ophthal-moplegia, Dieulafoy (4) reported 35 cases of sixth nerve palsy, 12 cases of third nerve palsy, five cases of fourth nerve palsy, and six cases of external ophthalmoplegia. The sixth cranial nerve was more often affected than the third one in two series (5,11). Conversely, in other series the third nerve is predominantly affected as the 14 patients reported by Weinstein and Dolger (12), included seven cases of third nerve palsy, six of sixth nerve involvement, and one with simultaneous involvement of both nerves. In an analysis of 811 cases of oculomotor palsies, diabetes accounted for 2.6% of third nerve palsy, 1.9% of sixth nerve palsy, and 0.6% of fourth nerve palsy (13). Finally, in Zorrilla and Kozak's series of 24 cases, 17 patients had an involvement of the third nerve, including two bilaterally, and seven cases of sixth nerve palsy, but no fourth nerve involvement (14).
In virtually all cases diabetic ophthalmoplegia occurs in patients with diabetes with more than 50 years of age, both in type 1 and type 2 diabetes. Rare cases have been reported in younger patients or even in children (15). The onset is rapid, within a day or two. In many cases, the patient experiences pains a few hours to a few days before noticing diplopia. Pain thus preceded the onset of diplopia in 14 out of the 25 patients reported by Green et al. (16) and in 18 out of the 22 episodes of oculomotor palsy that occurred in the 20 patients reported by Goldstein and Cogan (17). Pain seems common when the third cranial nerve is affected than when the sixth nerve is involved (14). Pain is usually aching behind or above the eye, and sometimes more diffuse, but always homolateral to the oculomotor palsy. Pain is often attributed to the involvement of the first and second divisions of the trigeminal nerve within the cavernous sinus (14), whereas others suggest a role for activation of pain-sensitive endings within the sheath of the third nerve as it traverses the cavernous sinus (8,18). Pain does not persist after the onset of diplopia.
Oculomotor dysfunction is often incomplete when the third nerve is involved, one or two muscles might only be paralyzed. In their series of 22 episodes of ophthalmoplegia observed in 20 patients, Goldstein and Cogan (17) mentioned 12 episodes of complete dysfunction, three episodes of nearly complete dysfunction, and three of partial paralysis. Ptosis is marked, the eye is deviated outward when the internal rectus muscle is affected; the patient is unable to move the eye medially, upward, or downward. Pupillary innervation is often spared, as in 75% of the cases in (17), whereas massive pupillar paralysis was observed in only two out of 20 patients. In another study (16) pupillary function was spared in 68% of cases, whereas Rucker (13) observed pupillary dysfunction in three out of 21 cases of third nerve palsy. Sparing of pupillary function permits differentiation of third nerve palsy of diabetic origin from third nerve palsy, resulting from compression of the nerve by an aneurysm of the posterior communicating artery in which pupillary dilatation is very common. The centrofascicular lesion found by Asbury and coworkers (8) at an autopsy of a patient with third nerve palsy accounts for sparing of pupillary function because of the relative sparing of pupillomo-tor fibers, which are peripherally placed in the third nerve (18). However, it has been suggested recently that isolated third nerve lesions in patients with diabetes, with or without pupillary sparing, could result from mesencephalic infarcts (19). In any case, brain magnetic resonance imaging should be performed to exclude a tumor, an aneurysm, or a hematoma.
Spontaneous complete recovery invariably occurs within an average 2-3 months, independently of the quality of control of hyperglycemia. Aberrant regeneration and synkinesis, which are so common after facial nerve palsy of different origin, do not disturb recovery of diabetic ophthalmoplegia.
Two serial section studies performed in patients with third cranial nerve palsy demonstrated a centrofascicular lesion of the nerve in its intracavernous portion (7,8). In the latter report, the axons were relatively spared on silver-stained sections. The myelin destructive lesion was 6-7 mm in length and the fibers placed at the periphery of the nerve trunk were relatively spared, which accounted for the pupillary sparing. The authors found no occluded vessel either intraneurally or in the nutrient vessels supplying the third nerve. In both reports the authors agreed that the observed centrofascicu-lar lesions of the third nerve were most likely ischemic in origin. However, it must be noted that nerve ischemia usually induce axonal nerve lesions, and not demyelinative ones. An inflammatory process of the type observed in biopsy specimens of the femoral nerve with partial ischemic lesions should also be considered.
Was this article helpful?